The document discusses how retailers and online businesses can improve customer recommendations and targeting by using data on users' online behaviors and interests across different sites and content, rather than just purchase histories. It provides examples of how understanding users' online interests in topics like hiking or cars could help recommend relevant products on other sites. The document also discusses how integrating online interest and behavior data with CRM systems can help target customers with tailored offers and develop product offerings to match evolving customer demand.
2. p.2
Content Discovery
Beyond retargeting and social
recommendation
Showing the same product (you
already bought?) can fail.
Product recommendation is not just for
retail sites.
Audience Data
Personalising all marketing and developing
offering with online purchase intent
Purchase histories fail for durable goods:
If you just bought a bike, you’re not buying a
new one soon.
4. p.4
Retargeting: “You looked at this, you will see it again”
Retargeting is better than no targeting, but is often annoying and ineffective.
Showing the same product too many times turns customers off, and if they already
bought it’s worse.
Social recommendation: “Users who bought this also bought..”
Pioneered by Amazon and very useful. But only works within the retail site, not
elsewhere.
Status quo in recommendation
5. p.5
1. Contextual targeting across sites and content types
Recommending products or messages related to the content user is viewing.
“You’re reading a bike review? Here are our offers on similar bikes”
2. Behavioral targeting across sites and content types
Recommending personally interesting products based on media viewing.
“You read about climate change and watched a video of a car show?
Here are our offers on hybrid and electric cars”
Other ways of recommendation
6. p.6
Iltalehti.fi newspaper is a top 2 site in Finland.
On the front page Leiki shows personalised recommendations
based on the user’s clicks throughout the media group’s network
with Leiki SmartPersonal. This provides easy access to personally
most interesting content.
Users see personal recommendations even on their first visit to
the site, if they’ve clicked on any other media group site.
On the article pages we help the user to stay engaged with the
whole media group with contextual recommendations from Leiki
SmartContext.
Case: Tabloid newspaper sends traffic
to premium content
Example: Car review with related content from various sites.
Recommendations include articles from (1) within the site, (2) from other media group sites and (3)
related car sale ads from online marketplaces.
7. p.7
Case: Multi-sector Retail Group
User reads an article on wild
mushroom hunting on
customer magazine. This
updates the personal
interest profile.
On the first visit to
department store site, the
front page recommends
products matching the user
profile, such as a GPS
navigator.
Contextual recommendations on the article page.
9. p.9
Dynamic creatives for Pre-targeting
Automatically scan web stores to promote the most relevant
items in dynamic creatives
Pre-target product offers in display ads both contextually and
personally
Each product can be presented alone or in a rotating multi-
product window
12. p.12
Define audiences around any interest
New types of segments that have been
unavailable for publishers before, e.g.;
• Interests on any product category
• B2B-decision makers - heavy readers of
advanced economy/business news
• Seasonal or temporary segments, such as Tyre-
Changers, Christmas shoppers, Eurovision Song
Contest
Market insight reports that summarize
semantic topics and trends on the
campaign
• Finding the right audience segments and topics
that most interests the advertiser’s customers.
Online audience targeting
Results
Interest targeting consistently
achieves a more than 200% increase
in CTR over standard targeting.
SmartSegment: Gigs & Events
Advertiser: Live Nation
SmartSegment: Beauty Queens
Advertiser: Nestle
SmartSegment: Video Games
Advertiser: Retailer Gigantti
SmartSegment: Sports Fanatics
Advertiser: Betting Agency ComeOn
1,00% CTR
0,66% CTR
0,43% CTR
0,41% CTR
13. p.13
Find interest differences between
demographic groups
Drill deep into specific topics or see
more general trend lines.
Compare the differences between
visitors from different geographic
areas or using different device types.
Discovering consumer interests
14. p.14
Drilling down into brand preferences
What are the interest differences
between consumers who choose
iPhones or iPads and those who
go with Android?
Analysis with select fashion
brands.
15. p.15
Daily goods
Somebody who buys fat-free milk every week will probably buy it next week
as well. Analyzing purchase histories works.
But what if they’ve been reading about oat milk recipes, but couldn’t find it
in your store?
Beyond purchase histories
Durable goods
Somebody who just bought a bike is not going to buy a new one next week.
Purchase history is not very useful.
Now they are reading office chair reviews. Who will market to them?
16. p.16
How to integrate online behavior?
Consumer purchase intent integration into CRM
1. Online interests are analyzed in detail, in real time
2. Online interests are mapped into your product taxonomy
3. Anonymous purchase intent data is sent to CRM
4. Anonymous cookie ID is sent to retailer on-site
5. Retailer maps anonymous ID to customer ID
17. p.17
Intent data
analyzer
CRM integration of online intent
Mapping into retail
product taxonomy
Retailer
site
Connection with
Customer ID
Analysis of
interests
from
browsing
Product purchase
intent transfer
Transfer of
anonymous ID
Media
Retailer
CRM
18. p.18
Using online behavior with CRM
Targeting
Make targeted offers of products related to consumers’ online interests.
In all channels!
Development of offering
Predict changes in consumer behavior before they are realized in purchases.
Develop your offering to match future needs of different demographics.
Stay ahead of your competitors who are looking at the past!
19. p.19
Using demographics with behavior
Start with interest, limit with demographic
Selling expensive sportcars? Good idea to combine interest in them with high
household income level.
Start with demographic, limit with interests
Tire-changers segment: People who own a car, but interests indicate that
they’d like someone else to change the tires. Such as busy parents with
children practicing sports.
20. p.20
Winning with online behavior
Offer what they want to buy before they’ve bought it
Send a personal paper / mobile coupon of a product with highest online
purchase intent to each loyal customer.
Find the product category with highest online intent and not yet realized
purchases in each locality. Tailor local shop offering before the competition.
Find the product segments with highest consumer interest that are missing
from your inventory – for each store.
21. p.21
Roy McDonald
President Leiki NA
+1 650 8676262
roy.mcdonald@leiki.com
Thank You!
Dr. Petrus Pennanen
CEO & Founder
+358 40 5020355
petrus.pennanen@leiki.com
Martin Säntti
Business Development
+358 400 977553
martin.santti@leiki.com
Jaakko Haarala
Business Development
+358 40 1384795
jaakko.haarala@leiki.com
Leiki HQ
Helsinki, Finland
Leiki NA
San Mateo, CA
www.leiki.com