2. AGENDA
On the subject of Behavioural Targeting,
Personalisation and Online Marketing
Optimisation.
Headline topics for the content are
1. Explain each concept
2. Highlight some goodbad examples
3. Describe how the Coremetrics applications can be
used to help our clients in this area.
3. USER CENTRIC APPROACH
Personalisation
Behavioural Online
Targeting Optimisation
CREATING EXPERIENCES THAT BUILD RELATIONSHIPS WHICH DRIVE RESULTS
4. Behavioural Targeting
Making the Online Experience Relevant
Behavioural advertising enables advertisers to reach valuable customers no matter
where they are surfing on the web. By targeting consumers behaviourally, advertisers
can engage people when they are most receptive to their message delivering these
with frequency and at scale.
Rob Blake, head of agency sales, AOL Advertising
We can attract prospects with customized campaigns according to their interests,
engage site visitors with dynamic content in response to their conduct and desires, and
put the right message in front of the right person at the right time. We can create a
more pleasant and more individual buying experience. We can quickly identify the
offers that will more likely convert those prospects to buyers.
Jim Sterne
GOALS: IMPROVING EFFECTIVENESS OF YOU
MARKETING STRATEGY
IMPROVING EFFICIENCY RE-TARGETING ADVERTS
Sources: IAB UK, AOL Advertising, eConsultancy, Wikipedia,
5. Behavioural Targeting
Making the Online Experience Relevant
IQOLA Case Study
The Idea:
Iqola was an invented energy drink – “the first drink to
improve your IQ” – to build awareness for a new brand
amongst a niche audience of senior media executives
purely via targeted display advertising.
The Campaign:
Spread over a three week period, the iqola campaign
re-targeted everyone who registered online to attend a
Microsoft Advertising organised conference on
behavioural advertising in late September 2009.
Impressions were delivered across AOL, Facebook,
MSN, Specific Media Network and Yahoo!
The Result:
50,000 impressions served.216 unique users.
67% of those polled recognised iqola as a brand they
had seen online.
Sources: IAB UK
6. Behavioural Targeting
Making the Online Experience Relevant
YAHOO PORTAL
Yahoo has a 2-petabyte, specially
built data warehouse, which it uses to
analyze the behavior of its half-billion
Web visitors per month, processing 24
billion events a day.
Yahoo! retains search requests for a
period of 13 months.
In response to European Regulators
Yahoo scrambles the last eight digits of
a users IP address after three months,
rendering them partially anonymous.
7. Personalisation
Making the Online Experience Relevant
Web personalisation is about delivering targeted content and adaptive web
experiences based on what you know about each visitor.
eConsultancy
Delivering customised content for the individual through web pages, e-mail or push
technology.
Dave Chaffey
Personalisation involves using technology to accommodate the differences between
individuals.
Wikipedia
Leveraging site visitor and customer profile data with marketing technology to create
relevant and target conversations between customers and merchandises.
Coremetrics
GOAL: IMPROVING EFFECTIVENESS OF YOU
MARKETING STRATEGY
Sources: eConsultancy, Wikipedia, Web Analytics Demystifies, Dave Chaffey
9. Personalisation
Making the Online Experience Relevant
1. Knowing your customer
Buying Decision Making
Process
a. New vs. Returning
b. Historical buying behavior
models
c. Immediate behavior
d. Sources
e. The Context
f. The dynamic environment
2. Proactively anticipating their
needs (Predictive Modeling)
3. Create “Real Time” Dynamic
individual profiles.
4. Automatically served them
the right product at the right
time using the right channel
using the right media format
Context – Relevant – Engaging – Automatic – Real Time – Enhancing Users Experience
10. Personalisation
Making the Online Experience Relevant
IMPLICIT PERSONALISATION EXAMPLE
1. Early adopter of
personalisation technology to
recommend products
2. Tracked/monitored user
behaviour as they navigate
the website (Registered user).
3. Use a rules-based filtering
model, based on "if this, then
that" rules processing, couple
with a collaborative filtering
approach, which serves
relevant material to
customers by combining their
own personal preferences
with the preferences of like-
minded others.
4. Email Re-Targeting model
11. Personalisation
Making the Online Experience Relevant
IMPLICIT PERSONALISATION – SEARCH EXAMPLE
Google uses factor weighting
model which includes as variables
factors as user history,
bookmarks, community
behaviour, site CRT and stickiness.
Need to be logged in Google
Account
Bing personalises search results
for all users based on an
individual's previous searches.
12. Personalisation
Making the Online Experience Relevant
EXPLICIT PERSONALISATION EXAMPLE
When a visitor elects a
profile that determines
they will see content in a
manner that it is specific
to their requirements.
Also known as
Customisation
This is something like recommendation
button shown along with the
each search result, by pressing this
button it will be shown in search
results to your friends(your
friends, chat and contacts in Google)
that you have “+1’d this” (that you like
or recommend this) if they have search
for the same term.
13. Personalisation
Making the Online Experience Relevant
IMPLICIT PERSONALISATION EXAMPLE
Personalised Subject Line: “I
coulda been a contender”).
Strong, Simple
Image: Timberland = boots –
sponsoring the Sundance Film
Festival in Park City = Snow
Simple Yet Provocative Headline
Content
Navigation, Pre-header, and
Mobile are all
14. ISSUES WITH PERSONALISATION & BT
Why some people are simply not interested
1. Anonymity preferred. Cookies deletion, log out.
2. Lack of relevance. People do not want a relationship with
companies that have no real relevance to them.
3. Lack of credibility. Relevant vs. Intrusion.
4. Lack of security. What about my personal details?
5. Technological barriers. Latest Mobile phone apps.
6. Infrequent contact.
7. Misunderstood terminology.
16. Personalisation
When things go wrong
AN EMAIL EXAMPLE – WRONG TARGETED SEGMENT
A “targeted” email campaign,
'Dedicate a Flower for
Mother's Day', supporting
the initiative with "send to a
friend" viral options and a
link to the Interflora Site for
information and purchase.
Mums received an email
from Interflora's epartner
Communicator, which
included a link to the page
where her message is
featured, and a confirmation
email is sent to the poster.
Badly targeted: It was sent
to everyone in their
mailing list
18. Behavioural targeting
When things go wrong
INTRUSIVE AND OUT OF CONTEXT
ADS
I made the mistake of
clicking an ad re those
websites once, more than 2
months ago.
They are still appearing in
disparate and not contextual
sites.
19.
20. ONLINE MARKETING OPTIMISATION
To capture and maintain your most valuable segment
Inaccurate Attribution Online
Marketing Model (last session or
last click 100% attribution)
1. Technological Barriers (first vs.
last touch strategy), vendors
limitations
2. Historical data not fully
available or too difficult
conciliating diverse data
sources
3. Segmentation based on
inaccurate Personas
4. High Risk in developing model
fitting the past but predicting
poorly the future
5. Lack of Resources/Budget
21. ONLINE MARKETING OPTIMISATION
To capture and maintain your most valuable segment
In Order to Boost Avg Order Value, Increase Conversion, strength loyalty and retention
we need to unlock the Jewel Box
1. Identify an universal key point, e.g.
Order ID – depending on business
objectives.
2. Conduct Path Analysis, Historical
Customer Purchasing Cycle, e.g.
last 3-6 months days data (proper
segmentation)
3. Click sequence analysis before
purchasing – timing depends on
product/industry. Use at least the
last known 4 touch points
4. Build a MKT attribution model
5. Predicting Model and automatic
services
6. Continually test your results
Econometrics↔ Multivariate Regression↔ (Math) Optimisation Model ↔ ?