Digital Marketing Measurement Framework - Martin Walsh

Making Sense of Data: Creating a
Structured Digital Measurement, Reporting,
Testing & Optimisation Framework.
Updated: Dec 2014
Martin Walsh
Director Digital Marketing

“If I had more time I would have
written a shorter letter presentation.”
Mark Twain

What I’ll try and cover
1. The Macro and Micro Context of Analytics – Slide 4
2. The Reality of a Marketers world – Slide 26
3. Introducing and operationalising a Digital Measurement Framework – Slide 32
• Revenue Performance Management via Marketing Automation – Slide 40
• Search Measurement – Slide 104
• Long Tail Advertising Performance Insights – Slide 120
• Social Influence (Media) Marketing Measurement – Slide 123
4. Qualitative vs Quantitative Analytics – Slide 133
5. Optimisation – Slide 169
6. Developing a Digital Taxonomy – Slide 175
7. Key takeaways – Slide 195

To be successful with analytics we must first
understand the Macro and Micro Environmental
Picture….

Senior executives now want businesses run on
data-driven decisions.*
These expectations can be met – but with a caveat.
For analytics-driven insights to be consumed – that is, to trigger
new actions across the organisation, they must be:
1. Closely linked to business strategy
2. Information must become easier to understand and act upon
3. Embedded into organisational processes to take action at the
right time
*Source: “Analytics: The New path to Value. MIT Sloan Management
Review & IBM Institute for Business Value Study 2010.

Each year marketers stake their reputations on
digital marketing programs and online activity.

And yet the marketplace is glutted with Digital
measurement tools and applications…..

So, why is there such a disconnect between
digital analytics tools and business results?

*Source: “Analytics: The New Path to Value.”
MIT Sloan and IBM 2010
3x 5.4x
Organisations that lead in
analytics outperform those who
are just beginning to adopt
analytics
Top Performers are more likely
to use an analytic approach over
intuition
Macro Environment: Key Finding 1
Good news: Analytics correlates to performance

Why Data Matters: Extracting Insights, Making
Better Decisions

Organisational, not data or financial concerns, are holding back
adoption

Organisation want to ‘see’ insights more clearly and act on them
Historic trend analysis
and forecasting
Standardised reporting
Data visualisation
Data visualisation
Simulation and scenario
development
Analytics applied within
business processes
Increased or
sustained value
Decreased in
value
Today 24 Months

Analytics use propagates across functions in a predictable
pattern
Brand and market management
General management
Workforce planning and allocation
Customer experience
Risk management
Product research and development
Customer Service
Strategy and business development
Sales and marketing
Operations and production
Financial management and budgeting
Transformed
Experienced
Aspirational

As adoption spreads, there is a growing demand for a greater
variety of skills and deeper expertise.
Centralised analytic units
LOB analytic units
At point-of-need
IT department
47%
71%
87%
Note: Chart combines responses to two questions. Percentage figures at top of bars reflect respondents
who answered “frequently” or “every day” to the question, “How often do you use information and
analytics to inform your actions and support decision making in your day-to-day role?” The height of
coloured segments within each bar reflect respondents’ answers to the question, “Where are analytics
primarily performed within your organization?”

Marketers Perceptions vs Reality
Most companies have not established a framework for customer
value and customer experience as key performance indicators .*
• Marketers overestimate the extent of their customer-centricity.
• Marketers overestimate their measurement capabilities.
• Marketers overestimate the scope of their customer segmentation.
*Source: “Integrated, Customer-Centric Marketing. Exposing Marketing’s Delusion.”
Forrester, August 2009
Micro Environment

Marketers Think They Are Customer Centric
Most organisations want to think that they put customers first,
but in reality most are structured around products, services and
line of business.
44%
35%
Customer Focused
Product/Brand Focused

Very few marketers use customer engagement as a primary
factor in their communications.
“We struggle to measure customer engagement. We don’t have a system to manage it, in part
because nobody has singular responsibility for managing it. Unfortunately, it’s just not a
priority.
Senior Marketer at a high-tech company
11%
32%
20%
Primary Factor
Often a Factor
Seldom or Never a Factor

Only half calculate customer value and fewer still, use it beyond
marketing.
• 50% say they have a customer value metric such as Lifetime
Value (LTV).
• 29% that have a value metric say it is not consistent throughout
marketing.
• 31% say that they are unable to measure incremental impact of
their marketing on customer value.
“We measure profitability where we can, but we are only starting to tackle what a customer lifetime
value metric might look like. Historically, our business lines have been too separate – the concept of
value for each was very different – but we are starting to bring those perspectives together and are
thinking about how to calculate lifetime value at a household level.”
A senior marketer at a financial services company.

A Siloed Approach is the Greatest Barrier to Adoption
of a Customer Management Approach

Marketers Believe They Consistently and
Systematically Measure Marketing
55% of marketers say that they use a standard or
consistent framework to measure all marketing
activity.
But…..

Measurement isn’t driving budget allocation
• 64% of marketers allocate budget across marketing disciplines
based on historical spending.
• 56% do so based on planned activity.
• Less than 30% use media mix modelling – which allows marketers
to understand the incremental impact of specific media and activity.
“We do try to track the impact of our marketing activity, but to a large extent we are winging it– we
make decisions based on past history, personal preferences, and what our agencies say is working for
other clients. Frankly, it’s a lot of trial, error, and guesswork.”
The head of marketing at a consumer packaged goods company.

Marketers Believe They Consistently and
Systematically Measure Marketing
Marketers put their emphasis on financial metrics.
• 63% of marketers selected frequency of purchases.
• 54% selected the frequency of interactions with their brand
• 47% selected the level and depth of brand interaction.
These are far more specific measures of engagement.
When asked how they assess the impact of their marketing
activity, 80% say financial impact – such as revenue or ROI.

Marketers are missing the point more than half the
time when calculating attribution for a transaction.

Marketers Claim Customer Segmentation
Extends Broadly
Mass, direct and digital communications aren’t integrated.
• 27% of marketers say that their marketing is integrated across brand,
direct and digital communications.
• 73% struggle to understand the impact that brand, mass, direct and digital
communication have on each other and therefore don’t plan programs
and campaigns to maximise their combined effectiveness.
“We compare data between brand and direct teams, but it’s not really integrated. Customers often tell
us that they see an offer in one medium and receive a different offer in another. Customer get
frustrated.”
A senior marketer at a financial services company.

Marketers Claim Customer Segmentation
Extends Broadly
Customer information isn’t integrated across marketing, sales
and service.
• Only 18% of marketers say that their activities are integrated – in that they
track and measure the impact each discipline has on the other.
• Only 30% say they unite disparate customer information from sales,
service and marketing into a single management framework for global
insights, targeting and measurement.
“One of our problems is that we don’t have single repository of customer information. It’s hard to
coordinate activity when each group has a very different view of the customer. It’s a big downside of our
siloed business structure.”
A senior marketer.

The Marketers Digital Challenge:
Measurement is More Important, But More Confusing!
Poor Marketers
42% of marketers are dissatisfied
with measurement and metrics!
- up 7% year on year.
Source: ANA/MMA study of 200 senior marketers from member companies
including IBM, Coca-Cola, Microsoft, ING and Nestle.

Yesterday
Evolving Digital Landscape Requires Evolving
Marketing Framework & Taxonomy
Today

• Variety of metrics and
methodologies to measure
effectiveness.
• Lack of standards for measuring
“effectiveness”
• Hinders ability to build digital
marketing business.
• Dizzying number of tactics.
• Will only get more complex
• Digital landscape can lead to
misguided and ineffective digital
marketing investments.
Navigating the Digital Marketing Chaos
Digital
TacticsCampaign Landings
Virtual Events
Digital Press Sites
Software Trials
Demos
In-Stream Videos
Rich Display Ads
Widgets
SEA
SEO
Live Webcasts
Mobile Codes
Social Networking
Blogs
SMS / Text
Promotional Emails
Digital
MeasurementVisits
Avg. Time on Site Site NSAT
Demos Begun
Trials Begun
Impressions
Cost per End Action
Install Base
Active Users
Subscribers
Fans
Comments
Event Registrations
CSAT
Average Duration
Response Rate
Churn Rate
Shares
Initial Responses
Search Driven Visits
End Actions
Organic Search Avg. Rank
?

Marketers Struggle With Measurement in Four
Fundamental Ways
1. Action without data
• High level goals without supporting
metrics to define them
• Measurement intended, but not
delivered (technology, budgets…)
4. Context without action
• No plans for improvement, continue as
before
• Actions “miss the window” to be
effective
2. Data without insight
• Unread reports (everyone gets 7+…)
• Data used to justify current activities vs
guiding new ones
3. Insight without context
• Confusion on which measures are
“right” and deserve focus
• Little proven business value from web
behaviours (“engagement”)

Key Micro Analytics Issues
Marketers and agencies are largely focused on:
1. The wrong metrics – too much focus on reach and not enough on engagement.
2. Drowning in data – looking at way too much data which is ultimately not insightful and actionable.
3. Not optimising activity / efforts against metrics.
4. No 360 degree view – marketers and agencies are looking at tactical metrics in isolation and not
looking at a complete 360 degree dashboard / picture of ALL their activities and importantly
customer behaviour.
5. Quantitative vs Qualitative – too much focus on quantitative data and not enough on qualitative
data, for example satisfaction.
6. No Taxonomy - not utilising a consistent and relevant digital marketing taxonomy.
7. No processes, no clear roles & responsibilities and a lack of skilled resources.
8. Over-engineering solutions – understanding what data you need to answer your questions, who
needs to access it and when and how your marketing applications need to interact with it is crucial
to implementing a simple solution.

Introducing and Operationalising a Digital
Measurement Framework

You Must First Understand the 90-10 Rule
of Analytics!
90% is about people…
(roles, responsibilities, frameworks,
processes, taxonomy, skills & resources)

You Must First Understand the 90-10 Rule
of Analytics!
…only 10% is about tools &
technology.

But wait! Before we dive into more detail we must first address a very common problem
in most marketing departments, ‘Silly Metrics’.
Too many organisations and too many marketers are using what I call ‘silly’ metrics to
‘measure’ marketing and in most cases have also inadvertently created a tsunami of un-
actionable reports.
• There are generally two categories of ‘silly’ marketing metrics:
Silly’ Marketing Metrics

“New path to value” is a five-point approach to
operationalising analytics
1
2
34
5
Focus on the biggest and
highest value opportunities
Within each opportunity,
start with questions, not
data
Embed insights to drive
actions and deliver value
Keep existing capabilities
while adding new ones
Use an information
agenda to plan for the
future

A Process for Inspiring Change
PADIE (Process-Application-Data-Insight-Embed) technique is a three-step process by which a company can
operationalize insights drawn from data: first, document processes and applications; second, use analytic
techniques to gain insight from data; and third, select the most appropriate ways to embed insight into operations.

Macro Environment
Within Each Opportunity, Start with Questions, Not Data….
Traditionally, organisations are tempted to start by gathering all available data
before beginning their analysis.
This leads to an all-encompassing focus on data management – collecting, cleansing
and converting data – with little time or energy or resources to understand its
potential uses.

The Big Picture: Where to start?
Pick your spot
1. Biggest and highest value opportunity
Prove the value
2. Start with questions 3. Embed insights
Continuous Value Delivery
4. Add capabilities 5. Information agenda

Awareness
Names
Prospects&Recycled
Marketing
Lead
Opportunity Sale
Nurturing Database
Inside Sales
Qualification SalesSalesLead
Recycled
Marketing
Revenue Performance Management
Marketing Automation Revenue Cycle Model - Example

Names
5,000
Prospects
1,000
*Active Prospect Database
250K
Marketing Leads
(300 + 7.5k) Nurture
Opportunities
3,744
Sales
1,498
Lead Type Variants
• Source
• Channel
• Industry
• Size
20% 70% (> 1 month)
30% (< 1 month)
15%
Inactive
(Last 6 months)
3.0% / month (7.5k)
75%
46% of prospects eventually become a
lead
40%
65% of all sales would follow this
model
80%
Awareness
100k
5%
*Cumulative Nurturing
Database
Sales Leads
4,680
60%
Revenue Cycle Model – Overview Example

Four KPI’s Summarise Marketing Impact
Marketing Generated
Revenue
% of sales revenue
uniquely created by
marketing
10% to
47%
Marketing Influenced
Revenue
% of sales revenue
touched by
marketing
30% to 82%
Investment-to-
Pipeline
Average cost of
demand creation of
sales revenue
1% to 4%
Investment-to-
Revenue
Average revenue
generated from $1
invested in demand
creation
$5 to
$20+
*Cross-industry averages
There are really only four marketing KPI’s or metrics you need to measure at an executive level
other than CSAT (Customer Satisfaction) or NSAT (Net Customer Satisfaction):
Real ROMI (Return of Marketing Investment)

Six Metrics Diagnose Marketing Opportunity
Database
% database
with bad/
incomplete
records
25%+
Inquiries
Raw
responders or
hand raisers
(Awareness to
Names)
2-5%
Marketing
Qualified Leads
(MQLs)
Leads ready to
be qualified
(Contacts to
Leads)
4% to 8%
Sales Accepted
Leads
(SALs)
Leads accepted
by Sales
45% to
75%
Opportunities
(Sales Qualified
Leads - SQLs)
A lead that is
opportunity in
pipeline
45% to
60%
Opportunity-
to-Sale
Leads that have
closed
20% to
30%
*Cross-industry average conversion rates
There are only six marketing KPI’s or metrics you need to diagnose marketing opportunities
created across your organisation:

Closed Loop Analytics - Revenue Performance Management

Closed Loop Analytics - Channel performance & Velocity

Closed Loop Analytics - Single view of customer

A/B Testing - Email, Landing Page and Forms

Key topic areas:
• Balance
• Flow
• Conversion
• Velocity
Filter/Drill into
data, e.g. by
Program Type,
Business Unit,
Geography etc.
Trends over time
Measure key metrics
associated with each stage of
the marketing & sales funnel:
• Balance of leads in stage
• In flow
• Out flow
• Conversion rate
• Velocity (time in stage)
Then have the ability to filter
this information by campaign,
products, business unit,
demographics, geography etc.
Revenue Cycle Analytics – Flows, Velocity, Conversions

Marketing Automation - Lead Scoring
• Lead scoring ranks a prospect or lead’s level of
interest and sales readiness according to a pre-
determined scoring methodology.
• Marketing communications are customised to
match the prospects characteristics and degree of
sales-readiness.
• Every activity the prospects or lead does is given a
score.
• When the scores reach a pre-determined
threshold, the prospect or lead is deemed
Marketing Qualified and therefore sales-ready and
transferred to the sales team.

Marketing Automation - Lead Scoring

How Lead Scoring Works in an Automated Marketing Context

Content & Behavioural Scoring – Active vs Latent Behaviour
Content Scoring Attributes
Example of Active vs Latent Behaviour.
Two different prospects with same score
but with different behaviour and time.

Explicit & Implicit Data
Explicit: What a prospect tells us / what we know Implicit: What a prospect does

Content Scoring and Optimisation
Constantly reviewing what content works
throughout the funnel.
In this example, 95% of all opportunities
visited the ‘Pricing Details Page’.
You could look at all of your content and see
whether Leads, Opportunities or importantly
Conversions/Sales and viewing / engaging
with it and when.

Marketing Automation – Prospect / Lead Characteristics
• There are three Qualification Characteristics;
Demographics, BANT (Budget, Authority, Need,
Timing) and Behavioural Indicators.
• Characteristics are critical signifiers that
determine the sales-readiness of a prospect or
lead.

Is to ensure a common, base-level of
understanding of the metrics, technology, tools
and processes that support Measurement,
Reporting, Testing and Optimisation
ADigitalMeasurementFrameworkObjective
64

Your Framework Must Be Focused on Identifying Insights
Which Connect to their Desired Outcomes
Insights
Recommendations
Findings
Information
Data
Of most value because it offers new
ways of looking at markets that lead to
competitive advantage.
Summarises the implications of the
research for the business.
Selected information that is of interest,
but lacking in implications.
Informs the business about a market,
but no indication of relative importance
of different pieces of information.
Of little value because it is usually
difficult to understand and interpret on
its own.
Source: Mohan Sawhney, Kellogg School of Management

Measurement Framework
Principles
1. Uncover business objectives,
understand existing
measurement capabilities, and
define measurement strategy
& taxonomy
2. Tagging strategy that aligns to
measurement strategy,
implementation plans, testing &
QA
3. Understand data sources,
develop integration schema,
integrate sources and test
4. Understand organizational
reporting needs, determine best
delivery mechanisms, develop
information architecture,
automation
5. Training, governance,
integration into business
decision-making processes
6. Define optimization strategies,
setup alerting on primary KPIs,
enable agencies with rich data

Conceptual Digital Measurement Framework Outline
Customer Lifecycle
Perceptual Journey
of Customers
Unaware of
Capabilities
Need to
Understand Value
Want to Investigate
/ Evaluate
Ready to Talk to
Vendor
Plan to Deploy &
Develop Skills
Use & Extract Value
Connect to
Community to
Share & Engage
Awareness Interest / Research Consideration Purchase Intent Purchase & Deploy Use & Support
Engage, Advocate
or Detract
Customer Interest
/ Tasks
• Discover
• Learn About
• Trust
• Authority
• Product
Comparisons
• Evaluation/Trial
• Analysis
• Case Studies
• Whitepapers /
TCO Reports
• Pricing /
Licensing
• How to Buy
• Learn
• Stay Up To Date
• Downloads
• Troubleshoot
• Community
Contribution
• Forums
• Customer Support
Digital Strategy
Get Attention / Get
Found
Inspire / Inform Influence, Persuade, Engage, Convert Satisfy Retain / Energise
Measurement
Objective
Maximise Reach ,
Discoverability &
Initial Responses
Maximise
Engagement
Optimise
Engagement
Optimise
Conversions / End
Actions
Maximise Sales
Maximise Customer
Satisfaction
Up Sell / Cross Sell /
Retain
KPI’s
(Examples and not
Exhaustive)
Impressions Unique Visitors
Content
Engagement Rate
# Prospects /
Responders
# End Actions (EA)
Lifetime Value
(LTV)
Initial Response
Rate (IRR)
Continued Action
Rate (CAR)
Cost per Engaged
Visit
Cost Per End Action (CPEA)
Churn
Qualitative - CSAT (customer satisfaction), NSAT (net satisfaction), Sentiment
Landing/Entry Page
Engagement Rate
Share of Voice
Cost Per Initial
Response (CPIR)
Data Sources
Ad Serving / Tracking – DoubleClick / Atlas
Search Engine Marketing – Marin & Covario
Cost Per Visit that
Consumes
Web Analytics – Coremetrics, NetInsights, WebTrends, Omniture, Google Analytics
Customer Satisfaction – Comscore, Get Satisfaction. Social Media / Sentiment – TruCast, Radian6, Converseon, Nielsen
Lead Scoring
Leads / Win
Revenue / Sales
Marketing Automation ((Email, Leads Nurturing &Management) – Unica, Exact Target, Responsys CRM- Dynamics CRM, Salesforce.com
Cost per Prospect
Opt Out Rate
Post SalesCustomer Lifecycle (B2B)
• Education
• Research
• Benefits
• Value
• Help
• How To
• eLearning
• Q&A’s
© Copyright 2011 Martin Walsh
http://twitter.com/martinwalsh

ROMI (Return On Marketing Investment)
• It is critical that an organisation develop a framework
for measuring the contribution of marketing
investment to the organisation as well as it’s ROI.
• A ROMI framework ensures the whole organisation is
on the same page, measuring the same things the
same way at the right time with the same tools and
appropriate expertise.

ROMI (Return On Marketing Investment)
1. ROMI (Return on Marketing Investment) should be a company-
wide project to ensure that you maximize your commercial
marketing investment.
2. The project requires leadership commitment & guidance to drive
sustainable process, infrastructure, organizational, and cultural
growth. Company-wide means across Business Groups,
Segments, Sales, Finance, HR and Marketing in both CHQ & the
Field.
3. ROMI is the vehicle that brings together the executive team,
project managers, subject matter experts (SMEs),
communications, and change management required to transform
innovative, bold ideas into real solutions that are broadly adopted.
4. ROMI is the fusion of art and science in Marketing.

The End Vision for ROMI
• The end vision for ROMI is that your organisation has greater
visibility into the effectiveness of its marketing investment.
• The organisation must enable marketing to be more
accountable through objective-based measurement.
• The organisation must ensure that its marketing disciplines,
processes, metrics, tools and organizational structure are
better aligned so that it has the infrastructure and capability
in place to deliver great marketing.
• Underlying this entire vision must be a commitment from the
organisation to improve customer and partner satisfaction.

Traditional Marketing Models Fail to Model Complex Buying Paths
Traditionally, marketers modeled customers’ decisions as they progressed from awareness through consideration, preference, action,
and loyalty — through what is called the marketing funnel. The marketer’s job was to move people from the large end down to the
small end. But now it’s time for a rethink, as the funnel has outlived its usefulness as a metaphor. Face it: Marketers no longer dictate
the path people take, nor do they lead the dialogue. We must rethink the marketing funnel because:
• Complexity reigns in the middle of the funnel.
• The most valuable customer isn’t necessarily someone who buys a lot.
• Traditional media channels are weakening.
• Consumers force brand transparency.

• Complexity reigns in the middle of the funnel. Awareness is still important; you need to know
that a product or service exists in order to buy it. And the marketer’s endpoint is still a
transaction. But, in between, other factors such as recommendations from friends or family,
product reviews, and competitive alternatives described by peers influence individuals. The
funnel’s consideration, preference, and action stages ignore these forces that marketers don’t
control. Rather than a clean linear path, the real process looks more like a complex network of
detours, back alleys, alternate entry and exit points, external influences, and alternative
resources.
• The most valuable customer isn’t necessarily someone who buys a lot. In this socially charged
era in which peers influence each other as much as companies do, good customers can’t be
identified solely by their purchases. Companies also need to track individuals who influence
others to buy. For example, a customer who buys very little from you but always rates and
reviews what she buys can be just as valuable as someone who buys a lot — her reviews might
influence 100 other people to buy your product. Tracking only transactions and loyalty at the
end of the funnel misses this significant element of influence.
• Traditional media channels are weakening. Marketers continue to use mainstream media
messages to move consumers into a consideration frame of mind. But passive consumption of
media is waning. Individuals dismiss or ignore marketing messages in lieu of information
available from an ever-increasing number of resources, such as product review sites, message
boards, and online video.
Traditional Marketing Models Fail to Model Complex Buying Paths

Today’s Complex Buying Paths
Marketing complexity means that
traditional methods and metrics fail
to address and capture the whole
story. Online metrics like unique
visitors to a Web site, number of
pages viewed, and time spent per
page mimic offline media metrics of
reach and frequency. But traditional
marketing and traditional
measurement doesn’t address or
indicate the engagement of an
individual; they fail to address or
capture the sentiment, opinion, and
affinity a person has towards a
brand as manifested in ratings,
reviews, comments in blogs or
discussion forums, or the likelihood
to recommend to a friend.

Your Organisation Must Move Beyond Campaign-Centricity and Take a Customer-Centric Approach to
Digital Marketing:
Traditional Approach
Campaign
Performance
Campaign-Centric
Audience Channel
Overall
Audience
• Generic
Content
• Channel-
specific
adaptation
Best Practice Approach
Campaign
Performance
Business
Impact
Customer-Centric
Audience Multichannel
Persona 1
Persona 2
Persona 3
• Cross-
Channel
Optimization
• Relevant
Product
• Meet Indv.
Needs &
Interests
• Right Time &
Channel
Means of Getting There
Data Integration
Behaviors Descriptors
Ongoing Program
Optimization With
Analytics
User Research & Segmentation
Strategy
Shift
Marketing Leadership: Strategy

Digital Marketing: Some Quick Definitions
There is still genuine confusion about the simplest, most
foundational, parts of digital metrics / analytics across
organisations:
1. Business Objectives
2. Goals
3. Metrics
4. Key Performance Indicators
5. Targets
6. Dimensions
7. Segments

1. Business Objectives
90% of the failures in web analytics, the reasons companies are data rich and
information poor, is because they don't have DUMB objectives.
Or they have just one (DUMB) Macro Conversion defined and completely ignore the
Micro Conversions and Economic Value.
DUMB:
• Doable.
• Understandable.
• Manageable.
• Beneficial.

2. Goals
Business objectives can be quite strategic and high level. Sell more stuff. Create happy
customers. Improve marketing effectiveness.
Goals are the next level drill down. It goes something like this. . .
Sell more stuff really means we have to:
1. do x
2. improve y
3. reduce z
The beauty of goals is that they reflect specific strategies. They are really DUMB. They
are priorities. They are actually things almost everyone in the company will understand
as soon as you say them.

3. Metrics
A metric is a number. That is the simplest way to think about it.
Technically a metric can be a Count (a total) or a Ratio (a division of one number by
another).
• Examples of metrics that are a Count is Visits or Pageviews.
• Examples of a Ratio is Conversion Rate (a quantitative metric) or Task Completion
Rate (a qualitative metric).

4. Key Performance Indicators
Key performance indicators (KPI's) are metrics. But not normal metrics. They are our
BFF's.
A key performance indicator (KPI) is a metric that helps you understand how you are
doing against your objectives.
That last word – objectives – is critical to something being called a KPI, which is also
why KPI's tend to be unique to each company.
• Business objective: Sell Stuff. KPI: Average Order Size.

5. Targets
Targets are numerical values you have pre-determined as indicators success or failure.
It is rare, even with the best intentions, that you'll create targets for all the metrics
you'll report on. Yet it is critical that you create targets for each web analytics key
performance indicator.
Best Buy: my KPI is still Average Order Value. But how do I know what's good or bad? If
I don’t have targets?

6. Dimensions
A dimension is, typically, an attribute of the Visitor to your website.
Here's a simplistic pictorial representation. . .
The source that someone came from (referring urls, campaigns, countries etc) is a
dimension in your web analytics data.

7. Segments
A segment contains a group of rows from one or more dimensions.
In aggregate almost all data is useless (like # of Visits). The best way to find insights is to
segment the data using one or more dimensions (like # of Visits from: Australia, USA,
UK, India as a % of All Visits).
You segment by dimensions and report by metrics.

Most marketers focus largely on:
1. Measuring Reach and Response
2. Measuring Sales
3. Not enough measuring Engagement, End Actions, Conversion and Velocity and
looking at it end-to-end
Delivered Impressions, Page Views, Unique Visitors, Win / Conversions – high-
level measures. But they are not very insightful in isolation and without looking
at engagement.
End
Action
MQPDelivered
Impressions
Initial
Response Opportunity
Initial
Response
1,000
Potential
Customers
Delivered
Impressions
5 Customers
Marketing Prospect Wins / ConversionEnd
Actions
MQPSuccess
Events
(Engagement)
Measurement – What Activities Should You Measure?
Raw Prospect Potential MQP
Creating Awareness & Engagement Generating Leads Sales
Execution of Marketing Tactics Pre-qualify & Route Responses Qualify & Close Sales
Opportunities
1 2
3

Key Measurement Elements – A Basic Scenario
Process Steps
1. Delivered Impressions
2. The # of people who click on ANY of the banner ads = Initial Responses
3. The # of people who watch the video = Engagement
5. Siebel measures MQLs, Opptys & Wins for each Ad/End Action combo
MQL
Opportunity
Win
5
Initial
Response
2
Landing Page
End Action
4
Data Form
Free White Paper
Delivered
Impressions
1
Banner Ads
Banner 1
Banner 2
Banner 3
4. The # of people who complete the Registration Wizard = End Actions
Engagement
Landing Page
3

Simplify Top-Of-Funnel Measurement
Customer arrives at site with
a need.
Visitors
Customer navigates site to
fulfill that need. (Engagement)
Success Events
Customer fulfills a need.
(Conversion)
End Action
1. Reach
2. Engagement
3. Call to
Action /
Conversion

Measurement – High Value Activities
2. Engagement
(Success Events)
3. End Actions
(Conversions)
1. Reach

Email Online Ads ATLSEOPaid Search Social Media
Primary Online Media KPI’s
• Delivered Impressions (DI) – 1,200,000
• Initial Responses (IR) – 21,450
• Initial Response Rate (IRR) – 1.79%
• Cost Per Initial Response (CPIR) - $2.96
• End Action Engagement Rate – 2.95%
• Cost Per End Action (CPEA) - $96.40
Primary Website KPI’s
• Unique Visitors – 78,465
• Engagement Rate (of Page) – 16.7%
• Continued Action Rate (CAR) – 3.5%
• Success Events – 7,860
• End Actions – 1,850
• End Action Conversion Rate - 2.35%
What Should We Measure – Top of the Funnel?

What Top of Funnel KPI’s?
You need to focus on KPI’s which can help you determine the effectiveness and
performance of your activity and which levers (tactics / channels) you can pull to
maximise ROI.
Example (purely illustrative):
Email Display Paid Search SEO ATL Social Media
Media Cost $977 $51,850 $70,618 N/A $257,600 $11,025
Impressions 100,000 1,481,403 2,250,000 2,500,000 11,500,000 845,600
IRR 4.25% 0.15% 0.52% 1.50% 0.85% 2.20%
CPIR $0.35 $77.80 $1.25 N/A $21.50 $1.50
Unique Visitors 4,250 2,222 11,700 15,000 97,750 18,603
Engagement Rate* 16.50% 12.78% 20.34% 27.20% 14.25% 27.20%
CAR 8.25% 6.24% 5.63% 9.97% 3.56% 11.13%
End Actions 399 255 983 840 2,346 2,100
End Action Conversion
Rate
9.4% 11.5% 8.4% 5.6% 2.4% 11.2%
Cost Per End Action $2.45 $203.33 $71.84 N/A $109.80 $5.25
*You would principally measure the Landing Page engagement rate and the Engagement Rate of the top Entry Pages
IRR = Initial Response Rate
CPIR = Cost Per Initial Response
CAR = Continued Action Rate
EA = End Actions
CPEA = Cost Per End Action

Why Focus on These Top of Funnel KPI’s?
By viewing this type of data we get focused on what is working and what isn’t working but in a
way which is actionable.
For example, if I you were simply looking at Site Visitors, Page Views and Site Clicks how would
you determine a poorly performing Landing Page? You could look at the bounce rate for the
Landing Page but this will not tell you what content on the Landing Page itself is driving
Engagement.
An overall Engagement Rate for a Landing Page or subsequent pages tells you how effective the
page is at Engaging a Site Visitor. It also allows you to optimise what is and isn’t working on the
page and it is a nice simple KPI for marketers and management to understand.
Example:
• 1% of Visitors download a whitepaper
• 10% of Visitors subscribe to an email newsletter
• 4% of Visitors downloaded a software trial
• 2.5% of Visitors subscribed to RSS feed
17.5% of Visitors engaged with content on the page / 4 engagement indicators =
4.4% page engagement rate

Why Focus on These Top of Funnel KPI’s?
End Actions are the last step in the process of meeting the customers need. Or, from the marketer’s
perspective, End Actions are when the customer has completed a Call to Action.
Generally speaking there is only ever one End Action per campaign or website. This is the single
customer need you are trying to fulfil or Call to Action.
For example:
If your campaign is about generating Awareness then an End Action could simply be watching a
designated video on the campaign site.
If the campaign is about generating Leads then the End Action could be a completed Registration as part
of an Online Assessment or even the successful initiation of a Click to Call or Click to Chat etc.
It is much easier to determine the End Action for a campaign than it is to determine for a persistent site
such as an Audience Relationship portal, e.g. Business Insights. However, it is important to have Success
Events and some form of End Action so you can measure the effectiveness of the site to meet these
Engagement KPI’s and importantly for you to optimise against. A good End Action for an Audience
Relationship portal would be signing up for an eNewsletter or membership to the site (if that exists.) If
the site was a product site then the End Action typically would be downloading a trial version of the
software but once again this also depends on what phase of the marketing lifecycle your website is
focused on. An Awareness End Action for a new product launch could still be watching a video if no
software download was available.

Important Note - Weighted Attribution
One of the key problems across marketing all around the world right now is that far too much, or all
performance attribution of marketing campaigns, particularly where there are digital elements, is being
given to the First Click or Last Click.
And sadly, far too much performance attribution is being given to the Last Click which in many cases is
Paid Search (PPC) or Organic Search (SEO).
Here are just some examples of the problem:
• If someone types in a branded search related term such as Commonwealth Bank Home Loans, the
customer has already been exposed to something other than PPC or SEO but that ‘other’ marketing
or advertising is getting no performance attribution.
• 71% of Paid Search clicks are navigational. This indicates that almost three quarters of your paid
search is not bringing in new prospects – it is simply delivering people who are already actively
looking for you and who have previously been exposed to some other form of advertising or
marketing.
• Approximately 44% of people who have clicked on a Paid Search Ad have been previously exposed
to a banner ad. But, the banner ad gets no performance attribution at all because the prospect
didn’t click on it and in most cases the marketer or agency uses the Last Click to determine what
worked – in this case paid search. (For the record, my peers and I stopped using CTR – click through
rate – on banner ads some 4 years ago because it is not the right performance metric for the
display advertising medium).
Now, there is no magic technical bullet here and if any vendor comes to you and say they do, start
running in the opposite direction. Great weighted attribution is derived through a combination of
specialised skills at a specialised agency, technology but importantly a proven methodology.

Engagement: A New Perspective on Marketing
As outlined earlier, if the marketing funnel no longer accurately reflects what marketers can influence, why do
they still cling to it? Because they can measure it, which is reassuring, even if it no longer accurately reflects
the real buying process. And, of course, there are no useful alternatives.
We believe that marketers need a new approach to understanding customers and prospects. This new type of
measurement — engagement — encompasses the quantitative metrics of site visits and transactions, the
qualitative metrics of brand awareness and loyalty, and the fuzzy areas in the middle best characterized by
social media.
The Elements Of Engagement:
Engagement goes beyond reach and frequency to measure people’s real feelings about brands. It starts with
their own brand relationship and continues as they extend that relationship to other customers. As a
customer’s participation with a brand deepens from site use and purchases (involvement and interaction) to
affinity and championing (intimacy and influence), measuring and acting on engagement becomes more
critical to understanding customers’ intentions. The four parts of engagement build on each other to make a
holistic picture.
• Involvement
• Interaction
• Intimacy
• Influence

Putting it all together:
Now that you know what information to collect, can we provide you with a formula to measure engagement?
Well, no, because it's different for each audience / segment / campaign. But you can start by identifying the
key metrics that are useful to you and use them to choose and work with vendors of measurement
technology. The engagement profiles you track will be different depending on your marketing strategy. Here
are four recipes for measuring engagement, matched to company objectives:
• If your objective is to create awareness: Start as you always have, tracking existing online ads, promotions,
and TV commercials. To this, add tracking content contributed by brand advocates, such as product reviews
or comments in discussion forums. Measure the sentiment of that UGC, including venues you don't control
such as third-party product review sites. Additionally, track influence through customers' intent to
recommend the product to a friend by conducting a survey or witnessing actions on peer review sites like
Epinions and TripAdvisor. A good example: A large national financial institution used brand monitoring
service Biz360 to identify key message pickup from its Superbowl ad in both traditional and interactive
channels, and used that buzz measurement to evaluate the expenditure.
• If your objective is to drive transactions: Track involvement to identify how people use your site (page
views, navigation paths, etc.) and merge that information with the effect UGC has on others' purchases.
This way you can discover the content and actions that increase a customer's likelihood to buy. For
example, CompUSA uses Bazaarvoice's services to provide product ratings and reviews, then tracks the
effect reviews have on the buying behavior of readers. They've found that customers acquired via review-
related searches convert at a 60% higher rate than the average customer and spend 50% more per order.

Putting it all together (contd.):
• If your objective is to build brand preference: Track involvement on your own site as well as referrals from
other sites to identify the content people use to make decisions. Understanding how intimate customers
are with your brand and what increases intimacy will establish what it takes to convince a person to choose
you over a competitor. For example, a large auto manufacturer launched a new SUV, but sentiment in
discussion forums (measured using brand monitoring services) indicated that the off-road capabilities of
the vehicle were not believable. Then, using imagery of the SUV performing in off-road conditions, the
company was able to track a change in perception, which resulted in outselling its competitors that
quarter.
• If your objective is to increase loyalty: Track recurring purchases, positive commentary, and intent to
recommend. Analyze the characteristics and behaviors of known loyalists to determine what you should do
to motivate potential loyalists to make the leap. Monitor sentiment about products in forums and in blog
comments to track those actions that commonly encourage or discourage repeat purchase and respond
accordingly through incentives, product modifications, or services enhancements. For example, in 2006,
Mini Cooper decided to boost its community efforts in a year with no new product announcements to
strengthen loyalty and avoid being seen as a fad. Working with MotiveQuest, it used brand monitoring
services and confirmed that the emotion expressed by owners in the Mini community set it apart from
other vehicles. In addition, Mini Cooper was able to monitor the results of the brand's community efforts
and evolve future communications.

Engagement Enhances Customer Insight – Social CRM:
Measuring Social Media Engagement (Social CRM) is a new concept. So far there are few measurement
vendors providing completely unified services to help tie it all together, however some solutions are
appearing, for example the recent (July 2009) announcement of integrated analytics by WebTrends + Radian6
+ SalesforceCRM.
Starting small and remaining focused on a few objectives at a time (such as connecting User Generated
Content to increased purchases or linking sentiment to loyalty) allows you to identify the metrics that matter
to you most. Integrate your customer perspective across channels and campaigns by using a variety of online
and offline metrics, so you can calculate the value of new channels and identify efficient strategies for them.
And consider the payoff of your hard work. You'll be able to:
• Measure and learn from things you currently ignore: Qualitative metrics like feelings, affinity, and
sentiment are difficult to track. And when you can collect the information, it has historically been
complicated or impossible to use with other data. But social media makes it easier for customers to offer up
their opinion, and, as a result, all that data is sitting out there waiting to be mined. Companies should track
these metrics on a regular basis with brand monitoring services, partly to get insights that no survey would
ever detect, since they come right from the minds of the customers. For example, a kitchen hardware
manufacturer should track commentary in discussion forums about home renovation to get ideas about
product shortcomings, identify new or improved products, and track impact on actual sales.

Engagement Enhances Customer Insight (contd.):
• Identify customers who influence others to buy: A person who contributes content, such as a
product review or a video of the product in use, may be far more valuable than the average
purchaser. As a result, your idea of who you consider a good customer should become more
robust, and you should make tactical offers with finer control to drive the behavior of these
customers. For example, a sporting goods retailer should identify customers who make
significant product reviews that influence others' purchases, then initiate programs to
encourage those customers to contribute more content.
• Encourage behavior across multiple touch points — online and offline: The engagement
profile crosses channels. Once you establish engagement profiles based on online data, the
next step is to tie in data from offline sources, such as in-store transactions, phone or catalog
orders, and customer service calls. For example, a quick service restaurant should determine if
customers that are heavy users of its in-store loyalty program are more likely to provide
favorable sentiment in online discussion forums or possess an increased likelihood to
recommend the food to a friend.

Define Key Indicators of Engagement
What characteristics describe an engaged visitor to our sites?
For example:
1. Download a whitepaper
2. Watch a video
3. Subscribe to email newsletter
4. Send a link to a colleague
5. Comment on post
6. Subscribe to RSS
7. Looking up a business partner
8. Download software trial

Prioritisation – Not All Objectives & Calls to Action
are Equal
Marketing Objectives:
• Driving Deployment 10
• Generating Leads 10
• Creating Awareness & Changing Perceptions 5
• Building Communications & Relationships 5
• Optimising Opportunities 5
• Readying Partners 5
• Enabling Infrastructure & Customer Data 5

Prioritisation – Not All Objectives & Calls to Action
are Equal
Success Event / End Action (High Valued Actions / Calls to Action):
• Awareness 10
• Event - Webcast 10
• Paid Event (live) 12
• Free Event (live) 5
• Initiate Sales Contact 5
• Marketing Collateral Kit 5
• On Demand Video 5
• White Paper 5
• Online Training 12
• Partner Offer 12
• Software Trial 20
• Profile Update 20
• Subscription – Newsletter 12
• Product offer 12
• Virtual Lab / Demo 20

Simplified Metrics – Site / Page Engagement Rate
Use simple engagement scoring to quantify your site’s current level of
engagement at a broad level
All percentages are the % of visitors who did what we wanted:
• % downloaded a whitepaper
• % subscribed to an email newsletter
• % downloaded a software trial
• % subscribe to RSS feed
• % watched
Total and then divide by the number of indicators.

Site / Page Engagement Rate Example
• 1% downloaded a whitepaper
• 10% subscribed to an email newsletter
• 4% downloaded a software trial
• 2.5% subscribe to RSS feed
17.5% engaged visitors / 4 engagement indicators =
4.4% page engagement rate

Search Engine Marketing (SEM)
Sadly in the area of SEM there is still:
• A significant lack of understanding of good search engine
marketing principles and which tactic is appropriate
depending upon marketing goals and objectives
• Over attribution of overall campaign performance to the last
click (ie. paid search)
• Too many generalities – B2C SEM is not the same as B2B SEM
• Lack of focus on insightful and actionable KPI’s

SEO Decision Funnel
As a visitor refines
their search criteria,
there are certain
patterns that
emerge in their
queries.

SEM Strategy MIX
With so many tactical options available
to today's search marketer, it can
sometimes be difficult to determine
which approaches are best for
achieving specific goals. Should press
releases be optimized for increased
brand awareness? Or, will a PPC (Pay
per click) campaign achieve better
results?

Understanding Search Engine Marketing
71% of sponsored search clicks are navigational
• This indicates that almost three quarters of our sponsored search is not
bringing in new prospects – it is simply delivering people who are already
actively looking for our URL.
50% of search dollars = branded search
• This indicates that more than half or our sponsored search is being spent
on navigational clicks.
44% of people who click on paid search are exposed to display
ads prior to the last click.
• Yet the majority of agencies and marketers are applying 100% of their
campaign performance to the last click (ie. paid search).

Search Budgets vs Search Opportunities
Paid Search Organic Search
Budget Clicks Conversions Budget Clicks Conversions
2007 88% 1 1.5 10% 5.6 1
2008 88% 1 1.5 10% 5.6 1
2010 91% 1 1.5 9% 5.6 1
As illustrated below, there is a significant misalignment of marketing budgets
to the actual performance of SEM.
For every paid search click there are 5.6 organic clicks yet SEO receives less
than 10% of the SEM budget and in most cases effort and resources.

B2C SEM is not the same as B2B: ie. Tech Buyers Search Engine
Behaviour – Key Takeaways
• 97% of tech buyers start with search
• Less than 12% of tech buyers us a single word search phrase
• 80% skim full page of results before clicking, 38% would click on paid search
• Even though page 1 ranking generates the highest volumes only 6% stick to just the page 1 search results,
53% will go to page 2 or 3 in a search, 41% say will go to 4+ pages
• Tech buyers use comparison oriented phrases close to end buying stage- i.e. Avaya Cisco IP phone system
comparison
• 53% use search to discover vendor solutions they are not aware of
• 67% willing to click in link of vendor they are not familiar with
• 27% tech buyers use mobile search at least once per week. 12% say they use it every day.
• Blackberry was top mobile device used in mobile search.
• 74% sign up for at least 1 rss feed, 26% sign up for more than 11 rss feeds
• North American's have a higher aversion to sharing their contact information compared to Asia and Europe
• When filling out a form, 81% of tech buyers will give their accurate email, only 27% will give their accurate
phone number
• Asia pacific uses video in tech buying twice as much as Nth America and Europe.
• IT buyers love video but video is a very underutilised form of communication when selling IT. I cannot
emphasise enough the power of video and you don’t need to be too worried about the high production
value. Use it whenever you can for product demos and more.
• 50.8% are watching videos on manufacturers (IBM, Microsoft etc) websites for information gathering to
identify potential enterprise technology solutions are manufacturer websites. This overall high usage of
manufacturer websites speaks to the effectiveness of video as a marketing medium for technology marketers.

Search Engine Marketing Measurement
It’s easier for marketers and management to focus on and
understand more meaningful KPI’s with the right tools:
Tools like BrightEdge can automate
the regular ‘auditing’ of your
website / pages, provide detailed
reporting on your own and your
competitors keywords.
Most importantly BrightEdge
operationalises optimisation
recommendations along with
simple step by step instructions as
well as workflow.
BrightEdge allows your marketing
team to tame organic search
engine marketing (SEO).

SEM – High Level KPI’s
SEO
• Traffic by Search Engine
• Share of Voice
• Traffic by Keyword
• Keyword Ranking
• High Value Actions (Success Events & End Actions)
– % Engagement Rate
– # of End Actions
– Which keywords are driving each HVA
– Which search engines are driving each HVA
SEA (paid search)
• IR: Initial Response
• IRR: % Initial Response Rate
• CPIR: Cost Per Initial Response
• ER: Engagement Rate
• CAR: Continued Action Rate
• EA: # End Actions
• End Action Conversion Rate %
• CPEA: $ Cost Per End Action

Search - Navigational
Separate out branded keywords when looking at clicks, cost and
performance.
• This will help you understand how paid search is contributing to our
overall advertising efforts. If branded keywords are driving a lot of sales, it
is likely that other marketing channels are generating interest in our
brand. In addition, analysing repeat-visit behaviour will help you optimize
your search campaigns. A high incidence of repeat visits is an indicator
that a keyword is being used navigationally. This information can be useful
in planning natural search optimization and could change how keyword
performance is being valued.

Test what happens if branded key phrases are bid down or
removed altogether.
• Several of advertisers in various studies were not bidding on any branded
key phrases. In most cases, those advertisers ranked at the top of the
natural listings for brand related key phrases, which is often true for
advertisers. A well-placed natural listing may be all the navigational
searcher needs to click-through and convert.

Consider all touch points that led to a conversion, not just the
last ad.
• As discussed previously, navigational search can mask the impact of other
advertising media due to last click attribution of conversions. In order to
accurately understand how different channels contribute to conversions,
use one tracking mechanism to measure the performance of all ad types
and look beyond the last ad when attributing value.

B2C and B2B Long Tail SEO
Short / Head tail terms
• Generally largest % of paid search is
invested in Head Terms
• Largest focus in terms of SEO
• Can drive high traffic volumes
• But can deliver low volume of leads as
traffic is generic and unqualified
• High cost due to high competition
Long tail terms
• Generally lowest % of paid search is
invested in Head Terms
• Lowest focus of SEO
• Will drive low traffic volumes
• But can deliver high volume of leads as
traffic is targeted and qualified
• Low cost due to low competition
Short / Head Tail = only 30% of leads opportunities
(but 85+% of budgets & SEO is spent here)
Long Tail = 70% of leads opportunities
(but 12% of budgets & SEO is spent here)
Marketers should also be spending more time on the B2B Long Tail:

Long Tail – KPI’s
Brand to non-Brand Ratio
• This is the % of your natural search traffic that comes from brand
keywords versus non-brand keywords. If the ratio is high and most
of your traffic is coming from searches for your brands, this signals
that your SEO is fundamentally broken. The lower the ratio, the
more of the long tail of natural search you are likely capturing. This
metric is an excellent gauge of the success of your optimisation
initiatives.
Unique crawled URL’s
• This is the number of unique (non duplicate) web pages crawled by
search engine spiders. Your website is like your virtual sales force,
bringing in prospects from the search engines. Think of each unique
page as one of your virtual salespeople. The more unique pages you
have, the more opportunities you have to sell through search
engines.

Search Visitors per Contributing Page
• This is the % of unique pages that yield search delivered traffic in a given
month. This ratio essentially is a key driver of the length of your long tail of
natural search. The more pages yielding traffic from search engines, the
healthier your SEO program is. If only a small portion of your website is
delivering searchers to your door, most of your pages – your virtual
salespeople – are warming the bench instead of working hard for you. You
can think of these as non-performing pages.
Keywords per Page
• This is the average number of keywords each page (minus the freeloaders)
yields in a given month. Put another way, it is the ratio of keywords to
pages yielding search traffic. The higher your keyword yield, the more of
the long tail of natural search your site will capture. In other words, the
more keywords each yielding page attracts or targets, the longer your tail.
So, an average of 8 search terms per page indicates pages with much
broader appeal to the engines than, say 3 search terms per page. The
average online retailer in a Netconcepts Study on the long tail of natural
search had 2.4 keywords per page.

Search Visitors per Keyword
• This is the ratio of search engine delivered visitors to search terms.
This metrics indicates how much traffic each keyword drives and is
a function of your rankings in the SERPs. Put another way, this
metric determines the height or thickness of your long tail. The
average merchant in the aforementioned Netconcepts study
obtained 1.9 visitors per keyword.
Index to Crawl ratio
• This is the ratio of pages indexed to unique crawled pages. If a
page gets crawled that doesn’t guarantee it will show up in a
search engines index. A low ratio can mean your site doesn’t
carry much weight in the search engines eyes.

Search Visitors per Crawled Page
• Calculated for each search engine separately, this is how much
traffic the engine delivers for every page it crawls. Each search
engine has a different audience size. This metric helps you fairly
compare the referral traffic you get from each. The Netconcepts
Study found that Bing and Yahoo! tended to crawl significantly
more pages, but the yield per crawled page from Google was
typically significantly higher.
Conclusion
• As you optimise your site through multiple iterations, watch
the aforementioned KPI’s to ensure that your heading in the
right direction. Those who are not privy to these metrics will
have a much harder time capturing the long tail of SEO.

Long Tail Advertising – CONTEXTWEB Study
CONTEXTWEB found the average click-
through rate for long-tail advertising during
the period* was 24% higher than for short-
tail advertising. All 20 advertiser verticals in
the study experienced click-through rate
gains, with the top category seeing a 50%
increase.
*The study, based on an analysis of digital media performance across more than
18,000 publisher sites and 1,000 campaigns from Q3 to Q4 2010
CTR Performance in the Short and Long Tail
Yet according to the CONTEXTWEB study, digital marketers continue to overspend on
the more costly short-tail without getting a significant boost in click-through rates to
justify the outlay. A couple of telling statistics from comScore cited by CONTEXTWEB:
while 70% of reportable display spending goes to short-tail websites , Internet users age
18 and over spend just 14% of their time online there.

Long Tail CTR Lift by Advertiser Vertical
All advertiser verticals saw increased CTRs in the long tail as compared to the short
tail. Advertisers in the “Alcohol” segment, for example, saw a 50% lift. The lowest
long-tail lift observed was 12% for auto advertisers.

Some other key findings detailed in the paper:
• All advertiser verticals saw increased click-through rates in the long
tail as compared to the short tail. Advertisers in the alcohol segment
saw the greatest lift, at 50%. The lowest long-tail lift was 12% for
auto advertisers.
• 76% of all content categories saw increased click-through rates on
long-tail sites, with the top three categories seeing a lift of 50% or
better.
• While 24% of content categories experienced a lift in click-through
rates in the short tail, ranging from 5% to 35%, these gains are often
offset by the price premium paid.
• Long-tail advertising provided significant unduplicated reach
compared to the short tail. In other words, advertisers can reach the
same type of audience with the same demographics as that of the
short tail, at a lower cost.
• CONTEXTWEB found select unduplicated reach to be at least 78% on
the long tail across 18 content categories, with “education” the
highest at 98%.
Click image to
download the study
(2.1MB .pdf)

Social Influence (Media) Marketing Metrics
“Social Influence Marketing is a commitment, not a campaign.”
Martin Walsh, Head of Digital Marketing, Microsoft – October 2007
“Social Networking is not Social Influence Marketing.”
Martin Walsh, Head of Digital Marketing, IBM – December 2009
Core Social Influence Marketing KPI’s (yours and your competitors):
• Topic Share of Voice
• Topic Comparison
• Sentiment
• Post Volume
• Author Volume
• Domain Volume
• Cost / Value per Impression and or Cost / Value per Engagement and or Cost / Value
per End Action

Social Influence (Media) Marketing (SIM) Metrics
SIM Measurement Framework
These are the three basic categories of Social Influence Marketing KPI’s:
Exposure Metrics
Engagement Metrics
Outcome
Metrics
Linked
What you measure depends entirely upon your marketing goals and objectives, ie. awareness, consideration, conversion. But, they are
linked in the sense that if you are measuring Outcomes (conversions) then you also have to measure Engagement and Exposures.

Social Influence Marketing Metrics
Here are two full sample reports (using real data) of what I consider best of
breed ‘Social Influence Marketing’ reports (both are from Visible Technologies
– Trucast):
Once again, data without analysis, insights and recommendations is practically
useless. You could use a myriad of tools and technologies but if you don’t have
the expertise and skills to transpose the data into actionable insights then you
are wasting your time.
(.PDF 3.4MB) (.PDF 4.2MB)
Click on images to download

Key Findings to Date
 Of sentimented data, 63.89% trend positively, 18.06% are mixed, and
18.06% trend negatively (fig.1). Positive sentiment has decreased by
8.38 percentage points, and negative sentiment has increased by 4
percentage points since the last reporting period.
 Of all posts relevant to the Windows topic, 16.8% contain some sort of
sentiment. This proportion of sentiment is high in comparison to other
technical topics.
 Post volume has remained quite steady since the last reporting period,
with minor increases in volume.
 When compared to other topics W7 remains the most highly discussed
topic.
Current Sentiment (fig.1)
Topic Comparison (fig.3)
Metrics Summary (fig.2)
Description
04/08/ –
04/21/
04/22/ –
05/04/
Change
Post
Volume
Total posts scored as
relevant to the topic
and date range.
3,330 3,428 +2.94%
Author
Volume
Total number of unique
authors for relevant
posts.
2,859 2,886 +0.94%
Domain
Volume
Unique domains that
hosted relevant posts.
1,001 1,073 +7.19%
Social Influence Marketing Metrics

Social Influence Marketing Metrics: Sentiment

Social Influence Marketing Metrics: Share of Voice

Social Media Influencer Ecosystem Mapping
It is important to first map your influencer ecosystem but remember that popularity (followers,
retweets etc) does not mean influence. Real influence is the ability to either directly or through a
third party influence someone to take action, drive conversations and or shift sentiment.

Exposure &Engagement Metrics
DeterminingROI/Value
In order to determine ROI / value of social media it is important to consider a
cost/benefit associated with the activity.
For example - Outreach and Engagement Activity, Cost, Value Report:

Bringing KPI’s together
It is one thing to measure some basic exposure and engagement metrics
such as Tweets, Facebook followers and other data points, but ultimately
if you are not measuring ‘outcomes’ – changing sentiment, driving a
webpage visit, driving a video view, driving online sales, driving Facebook
likes, event registrations or more (depending upon your marketing
objectives) then you are not really measuring the effectiveness and ROI of
your Social Influence Marketing efforts.
Example (purely illustrative):
Email Display Paid Search SEO ATL Social Media
Media Cost $977 $51,850 $70,618 N/A $257,600 $11,025
Impressions 100,000 1,481,403 2,250,000 2,500,000 11,500,000 845,600
IRR 4.25% 0.15% 0.52% 1.50% 0.85% 2.20%
CPIR $0.35 $77.80 $1.25 N/A $21.50 $1.50
Unique Visitors 4,250 2,222 11,700 15,000 97,750 18,603
Engagement Rate* 16.50% 12.78% 20.34% 27.20% 14.25% 27.20%
CAR 8.25% 6.24% 5.63% 9.97% 3.56% 11.13%
End Actions 399 255 983 840 2,346 2,100
End Action Conversion
Rate
9.4% 11.5% 8.4% 5.6% 2.4% 11.2%
Cost Per End Action $2.45 $203.33 $71.84 N/A $109.80 $5.25
*You would principally measure the Landing Page engagement rate and the Engagement Rate of the top Entry Pages
IRR = Initial Response Rate
CPIR = Cost Per Initial Response
CAR = Continued Action Rate
EA = End Actions
CPEA = Cost Per End Action

Ultimately It Must be About Actionable Insights
• Overall increase in positive sentiment, increase in
brand equity, greater event visibility and exposure
to non-search users.
• Volumes rose 200%+ around launch events, and
settled more positive than original positions.
Sentiment before, during and after launch events.
Heroes Happen Here Launch
 WOM had higher registration percentage
than any other campaign channel.
 Outperformed Paid Search by 2.5x.
 Simultaneous PR and advertising win.

Need More Emphasis on Qualitative vs Quantitative
Measurement
Quantitative
Qualitative

Need More Emphasis on Quantitative vs
Qualitative Measurement
Example:
I had a problem on a high profile website:
• Less than 32% of site visitors were successfully finding the content they were
looking for.
• Using our monthly Comscore Qualitative Site Satisfaction data we found out that
62% of people were coming to the site for Post-Sales Information but marketers
were focused solely on design, information and functionality for pre-sales! We had
a misalignment between customer needs and marketers expectations.
No amount of quantitative data would have provided this insight.

Why is Qualitative Measurement So Important?
• Helps you understand how satisfaction with all
aspects of the site tracks over time.
• Helps determine which aspects of the site
experience drive satisfaction.
• Allows you to identify actionable
improvements.
• Will reveal whether the sites are attracting
your target audience.
• Will reveal whether the sites are meeting the
needs of the target audience groups.
• Lets you explore whether site satisfaction and
brand preferences are correlated.

Primary Qualitative Metrics
• Overall Satisfaction - NSAT = VSAT – DSAT + 100
• Visitor Success Rate – Whether visitor found the information
• Primary Reasons for Visiting – Lifecycle = Pre-Sales & Post-Sales
• Satisfaction with Site Design
• Satisfaction with Site Content
• Satisfaction with Site Functionality
• Drivers of Satisfaction
• Who’s Visiting the Site
• Impact of Site Experience on Brand Preferences, Revisit Intentions
and Likelihood to Recommend
• Recommended Improvements

Site Satisfaction – Austria
Overall Satisfaction with Company
n = 775
60%
71%
33%
25%
7%
4%
0% 20% 40% 60% 80% 100%
Is a company I
trust
Is a
technologically
innovative
company
Agree (rated 8-9) Neutral (rated 5-7) Disagree (rated 1-4)
Attitudes towards Company
48%
57%
42%
35%
10%
8%
0% 20% 40% 60% 80% 100%
Security of
products
Ease of doing
business with
Company
Outstanding (rated 8-9) Neutral (rated 5-7) Poor (rated 1-4)
n = 756-764
Company’s Performance Attributes
48%
42%
10%
0%
20%
40%
60%
80%
100%
VSAT SSAT DSAT

Key Satisfaction Metrics – Austria
47% 44%
9%
0%
20%
40%
60%
80%
100%
VSAT SSAT DSAT
Overall Satisfaction with the Web Site
n =770
Performance Measures (n=276-766)
Outstanding
(8-9)
Poor
(1-4)
Quality of German translations 60% 5%
Quality of information provided in German 54% 7%
Usefulness of Information 53% 5%
Accuracy of Information 51% 7%
Topics Covered 50% 5%
Availability of information provided in
German
48% 9%
Depth of Information 46% 7%
Loading speed compared to other sites 56% 5%
Overall appearance 51% 5%
Organization of information 41% 9%
Ability to easily find information 34% 16%
Amount of time it took to find information 33% 16%
Site navigation 50% 8%
Search on the site 45% 11%
9%
17%
27%
47%
0% 20% 40% 60% 80% 100%
Don't
know
No
I found
some, but
not all
Yes
Whether Found Information
Looking For
9%
38%
53%
0% 20% 40% 60% 80% 100%
Worse
Neutral
Better
Compared to Other Technology Sites
n =768
n =661
ContentFunctionalityTools

Visitor Profile – Austria
IT Pro,
14%
Home
User,
40%
Dev.,
8%
Info.
Worker,
22%
BDM,
16%
Taxonomy
n =769
42%
4%
15%
13%
25%
0% 10% 20% 30% 40% 50%
Not asked/ Did
not answer
Don't know
Enterprise
(500+ PCs)
Mid-Market
(25-499 PCs)
Small Business
(1-24 PCs)
Organization Size
7%
13%
11%
22%
12%
21%
6%
8%
0% 10% 20% 30% 40% 50%
This is my first visit
Less than once a…
Once a month
A few times a month
Once a week
A few times a week
Once a day
More than once a day
Frequency of Visiting Web Site
Customer Lifecycle
Frequent
47%
Infrequent
46%
14%
19%
13%
16%
7%
16%
15%
0% 10% 20% 30% 40% 50%
Other
Support
Use
Deploym…
Purchase
Evaluation
Awareness
n =785
n =772
n =783
The highest
% in EMEA

60%
68%
93%
24%
18%
5%
16%
14%
2%
0% 20% 40% 60% 80% 100%
Buy a Company
product within next
6 months
Recommend the
Web site to
someone else
Re-visit the Web
site in the future
Would (rated 4-5) Neutral (rated 3) Would Not (rated 1-2)
Evaluation & Impact of Web Site Visit – Austria
n =721-758
Likelihood of Actions
47% 44%
9%
0%
20%
40%
60%
80%
100%
VSAT SSAT DSAT
Overall Satisfaction with the Web Site
n =770
3%
13%
70%
14%
0% 20% 40% 60% 80% 100%
Less than 20% (Find more than 80%
of materials on U.S. English site)
20% to 49%
50% to 99%
100% (Find all the information on
Austria Web site)
Proportion of Materials Found on Localized
Sub’s Network Performance Measures
(n= 746-757)
Outstanding
(8-9)
Poor
(1-4)
Quality of German
translations
60% 5%
Quality of information
provided in German
54% 7%
Usefulness of Information 53% 5%
Accuracy of Information 51% 7%
Topics Covered 50% 5%
Availability of information
provided in German
48% 9%
Depth of Information 46% 7%
n = 704

Top Drivers of Satisfaction for the Field
Potential
NSAT
Impact
H1 FYxx
Net
Score
∆ 6
Months
∆ 12
Months
∆ 18
Months
H1 FYxx
Driver
Rank
1.1 125 -9 -1 -6 21
0.7 113 -9 -1 -1 3
0.6 104 -3 +2 -2 18
0.6 77 -4 +6 +6 9
0.4 105 -16 -5 -8 2
0.4 109 -7 +3 +1 23
0.4 107 -4 +1 +1 16
0.3 93 -9 +1 -7 13
59% Somewhat Satisfied Swing Voters
To prevent the “At Risk” population
from moving to DSAT, prioritize efforts
to improve:
To move the “Leaning” population to
VSAT, prioritize efforts to improve:
•Demonstrates an understanding
•Helps me understand the software
roadmap
•Has appropriate frequency of
contact
•Demonstrates an understanding
•Products are easy to install •Is knowledgeable of tech solutions
H1 FY09
H1 FY08
DSAT
16%
DSAT
15%
At Risk
2%
At Risk
2%
Neutral
54%
Neutral
51%
Leaning
3%
Leaning
6%
VSAT
25%
VSAT
26%
IT Pro
Quality Cares Value Innovative Trust Security Ease Resources
NET
∆ 12
Months
NET
∆ 12
Months
NET
∆ 12
Months
NET
∆ 12
Months
NET
∆ 12
Months
NET
∆ 12
Months
NET
∆ 12
Months
NET
∆ 12
Months
Microsoft 104 -5 85 +2 73 -4 111 -4 103 +4 87 +2 79 -7 115 +2
IBM 144 +22 124 +26 121 +29 125 +9 133 +14 119 -10 114 +16 123 +17
Red Hat Linux 134 118 150 125 128 142 107 121
Oracle 104 102 81 96 105 111 100 107
Sun 128 115 121 132 120 131 109 121
Symantec 117 +4 101 +3 108 -5 115 +10 125 -2 129 -4 108 +9 101 -5
Qualitative Audience Performance Summary: Example Data Only
Swing Voter AnalysisPerformance on H1 FYxx Drivers
Competitive Analysis (Core Measures & Pillars)
H1 FYxx Top Improvement Areas
Lower price on products (incl. more
competitive pricing)
Licensing/improve licensing
Easier licensing policy
More reliable system (incl. less
system crashes)
Windows Vista Improved ease of use of products
Verbatim Distribution
49%
55%
46%
40%
4%
5%
5%
10%
15%
15%
10%
14%
22%
14%
20%
18%
9%
8%
17%
15%
H1 FY4
H2 FY3
H1 FY2
H2 FY1
Verbatim Themes Product (Net)
Security (Net)
Resources (Net)
Ease of Doing
Business (Net)
Cost (Net)
Scores showing significant improvement Scores showing significant decline Metric Leader
NSAT VSAT DSAT H1 FYxx
Base SizeH1 FYxx ∆ 6 Mos. ∆ 12 Mos. ∆ 18 Mos. H1 FYxx ∆ 6 Mos. ∆ 12 Mos. ∆ 18 Mos. H1 FYxx ∆ 6 Mos. ∆ 12 Mos. ∆ 18 Mos.
109 -8 -2 0 25 -2 -1 +3 16 +6 +1 +3 403
NSAT Performance
0.2%
0.2%
0.1%
0.4%
0.2%
0.2%
0.3%
0.2%
0.9%
0.5%
0.5%
0.2%
0.2%
0.2%
0.1%
0.1%
Products are relevant to my needs
Resources help me do my job better
Products are reliable
Products are secure from vulnerabilities
Products are easy to maintain
Resources help me get the most of our
products
Resources help me connect to the IT
community
Communicates business value
Share of impact moving DSAT to SSAT Share of impact moving SSAT to VSAT

Example: Qualitative Optimisation Priorities Example

General steps of a Measurement & Testing Approach:
Measuring Success: A Measurement Approach
1. Objective Setting
2. Measurement &
Testing Plan
3. Report Design,
Data Sources, SLA’s
5. Testing / eXp
6. Campaign
Monitoring
4. Setup &
Instrumentation
7. Reporting &
Optimisation
Campaign
Planning
Campaign
Execution
Campaign
Analysis

Establish Campaign Objective
• What are you trying to achieve?
• Example: Register customers for Power7 Server Launch
Events.
Set Clear Campaign Goal
• Check your benchmark tool for prior learning
• How will you know if you achieved your objective?
• Example: Register 1,000 customers in FY09.
Develop Measure, Test & Optimise Plan
• Identify the right KPI’s
• Define High Valued Actions (HVA’s)
• Choose the right tools for the job.
Develop Reporting Plan
• Determine Report / Dashboard Design
• Identify Data Sources
• Define SLA’s
Develop Instrumentation Plan
• Roles & Responsibilities
• SLA’s
• Setup/Instrumentation Checklists
MeasurementFundamentals

Determine
the purpose
of your site
Define what
you need to
track
Identify
primary
metrics
Create a
user flow
diagram for
tagging
Define
optimization
strategy
Set report
and
optimization
frequency
The6-stepMeasurement Plan

The type of site will determine the type of KPI’s /
metrics you will focus on:
– Audience Engagement Site
– Product Site
– Campaign Site
– Corporate Strategic Site
Step1:DefinethePurposeofYourSite

Key questions:
– What is the site/ campaign goal?
– What KPIs are critical for site / campaign optimization?
– Are these metrics trackable with our tools?
– If outcomes are not directly trackable, are there proxies?
• Example: MQPs/Opportunity Revenue as a proxy for Win Revenue
Step2:DefineWhatYouNeedtoTrack

• Primary KPI = Main site / campaign goal (End Action)
• Secondary KPIs = Metrics that optimization activities
will focus on (Success Events)
Optimizing secondary
metrics may improve
the primary metrics
Step3:IdentifyMetricsforSite/Campaign
Reporting/Optimization
Primary:
Event Registrations
Secondary:
Natural Search Referrals
Success Event Rates
Cost per End Action

Step4:CreateaUserFlowDiagram
andTagging Plan
The user flow is critical to get right and
understood by all as a basis for the tagging
and ultimately the measurement.

Tagging EnablesComprehensive Measurement
Initial Responses
•Banner/Search Clicks
•Opens
Success Events/
End Actions
•Clicks
•Profile Capture
Prospects
•MP/MQP #
•Cost per MP/MQP
Opportunities
•Oppty #
•Oppty Revenue
DoubleClick/NetInsights/Covario Siebel/Dashboard
Data
Source
Timing
Data available daily
2-3+ months required before valid
measurement

ATLAS/DoubleclickSpotlightTagging–WireframeExample
Event Landing Page – Best Practice
Red circles indicate
tracked Success
Events.

Developing andUsingCampaignCodesforTracking
What is a Smartcode?
“Intelligent” promotion code tied to every campaign and key campaign elements.
Carried through your systems for better campaign tracking.
Improved reporting that includes lead and pipeline data.
English Definition: The campaign code (e.g. Smartcode) is what enables
you to tie the cost and pipeline information to your campaign.

Marketers and the Agency should discuss which elements the agency is
optimizing and which aspects the company will optimize.
Step5:DefinetheOptimization Strategy

Need to provide consistent reporting; report design, determine roles &
responsibilities, data sources & SLA’s.
Step6:SetExpectationsonReportContent andFrequency
Intended
for Marketers
and
Management
Individual
campaign owner
Utilize for
planning bench
work comparison
formulation of
observations and
recommendations

ReportsandContext
It is important that we begin to look at digital data in a 360 degree
context of the different channels & tactics.
For example, looking at a paid search report alone and only looking at reach
(impressions) and initial response (CPC / CPIR) data is very misleading.
• 44% of all people who click on paid search results have been exposed
to display advertising prior to the click. Therefore it is wrong to
attribute 100% performance to paid search or the ‘last click’.
You need to look at paid and organic search data together and with broader
engagement/performance metrics.
More broadly you need to regularly look at data in overall context of a
campaign and or program and this also requires you to develop an overall
digital and or broader marketing scorecard for your organisation.

Why Data Matters: Context of Data Reveals
Answers

Example: Rolled Up Digital Marketing Management Scorecard & Metrics Definitions
KPI Measured by Definition
Discoverability Total SEO Health Audit Score This is the key Search Engine Optimisation (SEO) score to determine the overall discoverability and health of IBM.com online
Awareness and
Effectiveness of Paid,
Earned, Owned Media
Total $ Online Media Total amount of online media invested in the period.
Delivered Ad Impressions Number of successful targeted contacts or exposures for a marketing campaign.
Initial Responses Number of times a pre-defined initial or first action is performed by customers.
Cost Per Initial Response The actual cost of generating each Initial Response.
End Action Conversion Rate Number of end actions as a % of Initial Responses performed by customers who complete the desired advertising final action.
Cost per End Action Total cost of generating each End Action.
Total Leads Collected Marketing Prospects (SIPS)
Standard Individual Profile. A customer profiling standard that defines baseline attributes required for contacting individuals for the
organizations and determining the roles they play with their organizations.
Email Subscribers
Subscribers Total number of subscribers
New Subscribers Net number of new subscribers less un-subscribes
Opt Out Rate The % of opt out during the period
Open Rate The number of emails opened as a % of total emails sent
Initial Response Rate Number of times a click on a link within an email is performed as a % of the total emails opened
Engagement
Engagement Rate Effective rate of engagement between the visitor and content on landing pages and other primary pages
Success Events
Success Event is a pre-defined action performed by customers that indicates engagement with specific content on the page. Can help
determine ROI as not all customers will complete an End Action on a first or single visit.
End Actions
End Action is a pre-defined action performed by customers that completes the desired marketing campaign final action, response
mechanism, or call to action. This is typically the marketing end point. Some does not involve PII for instance, viewing a video.
Customer Satisfaction
VSAT
VSAT: % of respondents giving a very satisfied rating in response to a customer satisfaction survey. On a 4-point scale, 4 would be
a “very satisfied” score.
DSAT DSAT: % of respondents giving a dissatisfied rating in response to a customer satisfaction survey . On a 4-point scale, 1 and 2
would be a “dissatisfied” score.
NSAT NSAT: A calculated net satisfaction score. NSAT = VSAT – DSAT + 100.
Content Favourability
Growing industry trend to rate both our customer's satisfaction with specific content and more importantly, their intent to share the copy with
their peers (a measure of advocacy).
Reach of Site
Unique Visitors Number of unique individuals who visited your site during the report period.
Page Views Total pages viewed during the report period.

Context is King! You Need 360 Degree View of Data
Most marketers are looking at tactical metrics in isolation and
not using 360 degree views of all their activities.

Example: Integrated Digital Campaign Report

Example: Persistent Website Digital Report

Example: Digital Report
But, as outlined data without insights and recommendations is useless:

Benchmark Data is Crucial
Call-to-Action Audience= IT Pro
Paid Event 4.25%
Free Event 2.48%
Webcast 0.98%
EVAL /Trial 2.94%
Newsletter Subscription 7.00%
Kit 1.66%
Multiple 8.16%
Product Offer (free) 3.48%
Initial Response Rate Benchmarks by
Call to Action types
The best way to change behaviour, shift dollars to digital and execute the right digital
activity is to have a great measurement & reporting framework in place, capture historical
performance data and make it readily available to marketers and your agency.
Example:
As the adoption of standard KPI’s metrics, tools and processes grow, so does our learning:
Call to Action # of
campaigns
IRR Notes:
C-Suite Audience
Free Event 57 1.95% IT pros respond 3x higher to this value
proposition than do BDM
Product Offer 17 3.86% And this call to action is 2x better for this
audience…
…why are we using Events 3x more?
IT Pro Audience
Free Event 36 6.09% 6x more offers were driven with this CTA
but…
Trial CD/
Download
6 13.38%
…over 2x better IRR occurred with trials
offers
Are we leveraging the right Value Proposition for each Audience?

Benchmark Tool - Take a Snap Shot
• Snapshots are ready-to-run
reports that include a
summary and a graphical
representation of the data
• Perfect for a quick snap shot
of how a campaign performed
by:
– Call to action
– Customer group
– Geography
– Time period
– Other factors

Benchmark Tool - Using Snapshot Reports
• A marketer can use benchmark data to learn from previous
activities and investments and therefore more effectively plan
for the future.

Benchmark Tool - Customize a Query
• Run reports that target specific metrics:
– Time period
– Geography
– Segment
– Customer Group
• Compare the campaign to benchmark results
• Export results to Excel or Print

• Promotional Email IRR of 1.82% is below benchmark average, but failure rate
of 4.56% is above benchmark
Benchmark Tool - Customize a query to drill
down to the results for a specific country

Media Analytics Web Site Analytics
How many people were exposed to the
campaign message?
How many unique individuals visited
the website and how frequently do
they return?
Which media tactics are most cost
effective at driving conversion?
How much organic traffic does the
website get and where do they come
from?
Which Paid Search keywords drive the
most qualified visitors to the site?
What were the most popular paths
taken by visitors to the website and
where do they drop off?
Double
Click /
ATLAS
How well is your
ONLINE MEDIA
working to get
customers to the
site and do they
convert?
WebTrends
/ Omniture /
Core
Metrics/
NetInsights
How well is your
SITE attracting,
engaging and
converting
customers?
UsetheRightToolfortheJob

Data Sources Measurement Tools
Siebel
Atlas / DoubleClick
Marketing Automation
/ Google Analytics
BrightEdge / Covario
Marketing
Behavioural Database
Dashboard
ICE
comScore Survey
Behavioral
Attitudinal
TechnologyforMeasurement: DataSources&Tools
Overview
Integration
Note: Not intended to illustrate all data sources

Optimization
The following slides provide a guide for applying learning
from campaign measurement to optimize:
1. Marketing/Creative
2. Media
3. Web Site

MarketerOptimization
BasedonResults viaMeasurement
Optimization Frequency
Less
Monthly +
More
Auto
Program/Campaign
Source: Reporting/ROI
Communications Mix
Email/DM/Online/Events
Source: Reporting/ROI Analysis
Creative/Messaging/
Offer Strategy
Source: Testing
Creative/Messaging/
Offer Elements
Source: Auto-optimize Tools
Segmentation/Targeting
Primary Source: Testing
• Was the campaign or
program viable to
either expand or
repeat?
• Which channels or tactic
categories are contributing
the best ROI?
• Reallocate resources based
on results and feasibility.
• Auto-optimization tools
reallocate weight based on
the best performing
elements.
• Key elements include
creative components, offers,
and messages.

MediaOptimization
Less
Monthly +
More
Auto
Media Channels
Individual Media Tactics
Tactical Paid Search
Optimization
Content Sites
• Individual site performance.
• Placements on site, sections
within site
• Buying model (CPC vs. CPM)
Retargeting
• Performance by network
• UATs
Tactic Specific
Optimizations
(Examples)
• Daily optimization for ad
copy served & bids based on
dynamic bid, based
marketplace.
• Which media vehicles &
tactics are contributing to
most efficient & scalable
conversions contributing to
strong ROI?
• Shift resources based on
results.
Paid Search
• Keywords, ad copy, ad groups,
landing pages

WebSiteOptimization
Less
Monthly +
More
Auto
Campaign Wrap-Up Report
Source: Unica NetInsights, DoubleClick &
Covario
Monthly Site Reporting
Source: NetInsights, DoubleClick,
Covario
Calls to Action
Source: Testing
Dynamic Content
Delivery
Source: Auto-Optimization Tools
Campaign Landing
Strategy
• Comprehensive
campaign wrap-up
summarizes key
learnings that inform
strategy for future
campaigns.
• Overall site
design
• Flash /
Silverlight
issues (need for
2nd HTML
page?)
• Accessibility of
content (how
deep within site
CTAs are)
• Monthly site reporting
provides insight into overall
health of campaign.
• Example: Order
that CTAs appear
on the page
• Illustrates the impact
optimizations have had on
primary KPIs.
• Pinpoints areas of
opportunity for future
optimization.
• Auto-optimization tools based on
user preferences/profiles and/or
visitor behaviors
• Key elements can include creative
components, content, offers, and
messages.

MarketerOptimization
BasedonResults viaStatisticalTesting
Less
Monthly +
More
Auto
Program/Campaign
Source: Reporting/ROI
Communications Mix
Email/DM/Online/Events
Source: Reporting/ROI Analysis
Creative/Messaging/
Offer Strategy
Source: Testing
Creative/Messaging/
Offer Elements
Source: Auto-optimize Tools
Segmentation/Targeting
• Which target is more
responsive to a campaign or
campaign elements?
• Are top deciles responding
at expected rates?
• For strategic guidance to
inform new initiatives or
current campaign strategy,
testing controls outside
influences to isolate which
factors or combination of
factors improve results.

WebSiteOptimization
BasedonResults viaStatisticalTesting
Less
Monthly +
More
Auto
Campaign Wrap-Up Report
Covario
Monthly Site Reporting
Covario
Calls to Action
Source: Testing
Dynamic Content
Delivery
Source: Auto-Optimization Tools
Campaign Landing
Strategy
• Can we reduce abandonment rates on
the site by modifying existing landing
page design/layout?
• Site load time is often
addressed to reduce
abandonment rates.
• Items to test: design, layout, creative
elements, etc.
• Can we improve engagement
rates with key end actions by
providing clearer calls to
action?
• Items to test: messaging,
expanding on value
proposition, increasing
prominence of element, etc.

Digital Marketing Taxonomy
Goal
Define and deploy a digital marketing taxonomy to enable more
effective digital marketing.
Strategy
1. Establish a common set of definitions and success metrics for
the digital marketing “mix”.
2. Integrate taxonomy with relevant marketing tools and
processes.

Why Marketing Taxonomy Matters
Taxonomy
Marketer
Customer
A Marketing Taxonomy…
1. …provides a common language
to enable integration of tools and
processes.
2. …provides increased
organizational transparency
across planning, execution,
measurement and reporting.
3. …enables more productive and
effective marketer to customer
interactions via integrated
tools/process and increased
transparency.
4. …enables better BI through
standardized customer
responses.
5. …reflects the changing needs
and behaviors of our customers.
ProcessesTools
ReportingMeasurementExecutionPlanning
1
35
Response
2
4
Integration of Tools and Work Flow:

Level 1:
Tactic Families
Level 2:
Tactics Tactic Metrics
Digital Destinations
Digital Engagements
Digital Advertising
Digital Direct
Social Media
Digital Events
Mobile
Search
Organizing the Chaos with Taxonomy
Digital
TacticsCampaign Landings
Virtual Events
Digital Press Sites
Software Trials
Demos
In-Stream Videos
Rich Display Ads
Widgets
SEA
SEO
Live Webcasts
Mobile Codes
Social Networking
Blogs
SMS / Text
Promotional Emails
Digital
MeasurementVisits
Avg. Time on Site Site NSAT
Demos Begun
Trials Begun
Impressions
Cost per End Action
Install Base
Active Users
Subscribers
Fans
Comments
Event Registrations
CSAT
Average Duration
Response Rate
Churn Rate
Shares
Initial Responses
Search Driven Visits
End Actions
Organic Search Avg. Rank
• Measurement plans must be established
for each tactic – a task compounded by the
variety of metrics available.
• Moreover, these measurement plans are
often created in silos through ad-hoc
processes, resulting in limited ability to
draw conclusions across parallel efforts.
• Marketers must first select a digital
marketing tactic from the myriad of tactics
available.
• Often, limited guidance is available to
assist marketers in selection.
A digital marketing taxonomy
serves as a framework to
organize the digital “clutter” into a
construct that makes sense and
enables more effective digital
marketing.

Start With the Customer Perspective
Digital Engagements
Digital Destinations Digital Advertising
Digital Direct
Social Media
Mobile
Digital Events
Search

Different tactics, channels and even audiences require different
metrics depending upon your objectives.
1. Exposure
2. Engagement
3. Conversion
Reach vs Engagement vs Response

A Digital Interaction Mode is… A Digital Tactic is…
A family of distinct marketing tactics that share a unique
interaction experience with customers.
A mutually exclusive executable digital marketing
element that can be discretely measured. Tactics fit
cleanly and exclusively into one digital interaction
mode.
Level 1 Digital Interaction Modes Level 2 Digital Tactics
Digital Destinations Websites
Campaign Landing
Evergreen Site
Digital Press Site
Dark Site
Digital Engagements
Experiences customers engage with
on-demand and, often, anonymously
IM AnswerBot
Game
Demo
Software Trial
Syndicated & Embeddable Content
Video
Digital Advertising
“Controlled placement” digital
elements
Display - Static
Display - Animated
Display - Rich
In-Stream Video
Text Ad
Audio Ad
Custom Integration
Roadblock
Over the Page Unit
Digital Direct
Experiences that require some form of
customer opt-in and, often, mandate
refreshed content
Click-2-Call / Chat / Email
Podcast / Vodcast
Widget / Gadget / App
Loyalty Rewards Program
Contest / Sweepstakes
RSS
Newsletter
Promotional Email
Social Media
Experiences involving social interaction
or sharing between users
Social Networking
Blog / Microblog
Crowdsourcing
Media Sharing
Wiki & Collaboration Software
Social Bookmarking
Discussion Forum
Virtual World
Social Widget
Digital Events Time-boxed digital group interactions Live Webcast Virtual Event
Mobile
Mobile device-based digital
experiences
Mobile App
SMS / Text
Mobile Code
Location Based Service
Sponsored Mobile Content
Search Search SEA SEO
Organizing Principles: 2-Tier Taxonomy of
Interaction Modes & Digital Tactics

Digital Marketing Taxonomy
The Taxonomy…
1. …provides a common language for digital marketing efforts .
2. …creates a framework for the integration of tools and processes and the
implementation of best practices.
3. …enables better BI through standardised KPI measurement.

Taxonomy Flat File
The Taxonomy is backed by a detailed flat file that establishes standard metric
definitions, identifies data sources, and informs metric calculation for reporting.

Measurement, Reporting, Instrumentation
There are a number of key components required to
operationalise this capability:
1. Develop a Digital Measurement Framework
2. Develop a Digital Taxonomy
3. Develop a Digital Measurement Centre of Excellence
• Playbook
• Process
• Tasks
• Roles & responsibilities
• Resources / skills
• Documents
• Briefs
• Plans
• Reports
• Instrumentation Checklists / Instructions

Digital Marketing Measurement Framework - Martin Walsh

Digital Marketing Measurement Framework - Martin Walsh

Recommended

More Related Content

What's hot

What's hot (20)

Viewers also liked

Viewers also liked (20)

Similar to Digital Marketing Measurement Framework - Martin Walsh

Similar to Digital Marketing Measurement Framework - Martin Walsh (20)

More from Martin Walsh

More from Martin Walsh (13)

Recently uploaded

Recently uploaded (20)

Digital Marketing Measurement Framework - Martin Walsh

Editor's Notes