Three case studies of real-world web content personalization systems including ecommerce, breast cancer diagnosis and treatment information, and prescription drug formulary data
5. “…involves using technology to accommodate the
differences between individuals.”
“Delivering the right information, to the right people, at
the right time, in the right format and language.”
“The combination of a ‘person’ and a bunch of
‘alization’”.
6. Person alization
(Business or website (Machinery that measures,
owners, content writers analyzes, and eventually
and editors, UX folk. partially or fully automates
NOT the end user.) content delivery)
7. When not to personalize
When you have a crappy
website
When you don’t have a
UCD program in place
When you have a great
website (but there’s no
value add)
8. When you might consider personalization:
To surface relevant content that a user would otherwise not
know to look for
To identify high profit customers and then make them more
profitable
To partially or fully automate more effective marketing to
users or audience segments
Key: add value above and beyond an already high-performing
site or app
9. Key Steps of Personalization
Measure!
Traffic, user feedback, customer inquiries,
sales data, user demographics—anything
that will help you make measurable
improvements.
10. Key Steps of Personalization
Develop and Test
Develop and test these improvements. Start
simple. Collect data on what works. Refine.
11. Key Steps of Personalization
Automate
Find ways to automate delivery of personalized
content or application functionality to users—when
it makes sense.
18. Case Study 1
Breastcancer.org CA 15.3
Colloid
Thermography Cribriform
Subtrac(on Probe Technology Chromogenic In Situ Hybridiza(on
Mucinous
Oncogene Overexpression
19. Case Study 1
Breastcancer.org CA 15.3
ER/PR+ Digital Tomosynthesis
Colloid Papillary
Thermography Cribriform
CA125
Subtrac(on Probe Technology Chromogenic In Situ Hybridiza(on
Mucinous Ki‐67
Comedo Her2/neu
Oncogene Overexpression
Progesterone Receptors CA 27.29
20. Case Study 1
Breastcancer.org CA 15.3 Posi(ve margins
ER/PR+ Digital Tomosynthesis
Colloid Papillary
Thermography Cribriform
CA125
Ductal Lavage Ductal Carcinoma In Situ
Subtrac(on Probe Technology Chromogenic In Situ Hybridiza(on
Mucinous Ki‐67 ImmunoHistoChemistry
Comedo Her2/neu
Oncogene Overexpression
Progesterone Receptors CA 27.29
Fluorescence In Situ Hybridiza(on
21. Case Study 1
Breastcancer.org
And then, if you’re like an awful lot of folks, you’ll click on the first
organic result, which is breastcancer.org…
23. Case Study 1
Breastcancer.org
Lessons from User Research…
Clear navigation and powerful
search were not enough
But, oh, the risks of
personalization…
24. Case Study 1
Breastcancer.org
Risks of Personalization…
Overly negative and/or poorly targeted
research articles
Too much info about advanced new
treatments that they were “missing out on”
Users unable to report their own diagnosis
information accurately
25. Case Study 1
Breastcancer.org
Five vocabularies were developed to assign metadata to content.
Clinical
Audience Situa(on Perspec(ve Topic
Characteris(cs
26. Case Study 1
Breastcancer.org
Audience Patients
Family & Friends
Press & Public
Clinicians & Providers
Worried Well
27. Case Study 1
Breastcancer.org
Situation Just Diagnosed
Waiting for Test Results
Undergoing Treatment
Recovery & Renewal
Metastatic Cancer
End of Life
28. Case Study 1
Breastcancer.org
Clinical Characteristics Lobular Carcinoma In Situ
Ductal Carcinoma In Situ
Inflammatory Breast Cancer
Recurrent Cancer
Metastatic
Lymph Node Involvement
Male Breast Cancer
Post-menopausal
*This is a partial list
30. Case Study 1
Breastcancer.org
Topic Environmental Risk Factors
Genetic Risk Factors
Symptoms, Self-Detection & Breast Self-Examination
Medical Screening and Testing
Dealing with Cancer Fear
Surgery
Chemotherapy
Radiation Therapy
Hormonal Therapy
*This is a partial list
31. Case Study 1
Breastcancer.org
Audience Patients
Situa,on Undergoing Treatment
Clinical Lymph Nodes Removed: 20+
Characteris,cs
32. Case Study 1
Breastcancer.org
Audience Patients
Situa,on Undergoing Treatment
Clinical Estrogen Receptor+
Progesterone Receptor+
Characteris,cs Pre-menopausal
33. Case Study 1
Breastcancer.org
Audience Patients
Situa,on Undergoing Treatment
Audience
Clinical Estrogen Receptor+
Progesterone Receptor+
Characteris,cs Pre-menopausal
34. Case Study 1
Breastcancer.org
Metadata for
personal profiles
was made possible
by adding structure
to already in-use
signature
information
35. Case Study 1
Breastcancer.org
How it works
Push personalized content to
users based on profile info:
Email
Website content
RSS (in funding)
Breastcancer.org staff create
rules for content metadata and
personal metadata overlap
“fingerprints” that add value
36. Case Study 1
Breastcancer.org
Why Personalization is Important
Allows breast cancer patients to fully
understand their diagnosis—making them
more likely to question their doctors.
Provides extra information that users
would not necessarily know to look for via
search or navigation.
37. Case Study 2
Data Warehouse Client
Formulary data (drug coverage across states, health plans, and tiers) is populated by a
team of pharmacists that keep the database up-to-date daily via another piece of the
application.
38. Case Study 2
Data Warehouse Client
Why Personalize?
Pharmaceutical corporations do
their marketing everywhere.
To convince doctors that a drug
is cheaper than an alternative.
Pharmaceutical sales reps can
tailor printed marketing materials
to individual doctors.
39. Case Study 2
Data Warehouse Client
The “Person”
Pharmaceutical corporation
sales staff
40. Case Study 2
Data Warehouse Client
The “alization”
Custom ColdFusion
application
PostgreSQL database
iText Java Library PDF
rendering tool
41. Case Study 2
Data Warehouse Client
How it works
Sales reps start by selecting:
Drug Class
Drug
State
Health Plan
42. Case Study 2
Data Warehouse Client
How it works
Sales Reps may then select
options such as:
Whether to display co-pays
The order in which results
appear
Bolding of results
43. Case Study 2
Data Warehouse Client
How it works
Sales reps may add “personal
touches”, and then generate a print-
ready PDF file.
Automatically sent to approved print
vendors for production.
44. Case Study 2
Data Warehouse Client
Why Personalization is Important
Allows sales staff to quickly turn up-to-date,
national database of formulary information into
glossy, personalized marketing materials.
Fine control over data presentation,
geographic coverage, and formatting add to
effectiveness of marketing materials.
46. Case Study 3
ShipYourReptiles.com
Why Personalize?
To target customer service
resources, volume-based
discounts, and value-add features
towards certain purchasing profiles.
Convenience = increased profits
Preferred status = repeat business
and collection of valuable feedback
47. Case Study 3
ShipYourReptiles.com
The “Person”
ShipYourReptiles customer
service and sales staff:
Reptile shipping experts
Thorough understanding of
UPS rate structure
Capable of making account
upgrades
49. Case Study 3
ShipYourReptiles.com
How it works
Due to unique requirements of UPS
shipping label purchase, all users must
create accounts to buy labels.
Users may personalize their own accounts:
Stored payment methods
Address book for ship-from and ship-to
addresses
50. Case Study 3
ShipYourReptiles.com
How it works
Account managers may
review order activity, and
upgrade accounts for high-
volume customers.
51. Case Study 3
ShipYourReptiles.com
Why Personalization is Important
In eCommerce, every fraction of a percentage point counts
when multiplied across thousands of orders and customers,
day in and day out.
52. Conclusion: Review of Key Steps of Personalization:
Measure!
Develop and Test
Automate