Web Content Personalization: Three Case Studies

The Anatomy of a Personalization System: Three Case Studies

Derek Olson, Vice President, Foraker Design

Is your website acting like it’s on a first date instead of an anniversary dinner?

Case Studies:

1 2

Custom Web to Print

3

Snakes on a Plane

Improving Health Care Decision Making

What is

personalization?

“…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’”.

Person

alization

(Business or website owners, content writers and editors, UX folk. NOT the end user.)

(Machinery that measures, analyzes, and eventually partially or fully automates content delivery)

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)

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

Key Steps of Personalization

Measure! Traffic, user feedback, customer inquiries, sales data, user demographics—anything that will help you make measurable improvements.

Key Steps of Personalization

Develop and Test Develop and test these improvements. Start simple. Collect data on what works. Refine.

Key Steps of Personalization

Automate Find ways to automate delivery of personalized content or application functionality to users—when it makes sense.

Case Study 1

Breastcancer.org

Case Study 1

Breastcancer.org The “Person” Breastcancer.org editorial and medical advisory staff.

Case Study 1

Breastcancer.org The “alization”  

Custom Ruby on Rails

 

PostgreSQL database

 

SDL Tridion CMS

 

Google Analytics

Case Study 1

Breastcancer.org

“You have breast cancer...”

Case Study 1

Breastcancer.org

?

Case Study 1

Breastcancer.org

Subtrac(on
Probe
Technology







Chromogenic
In
Situ
Hybridiza(on


Case Study 1

Breastcancer.org

CA
15.3


Colloid


Thermography


Subtrac(on
Probe
Technology




Cribriform




Chromogenic
In
Situ
Hybridiza(on


Mucinous
 Oncogene
Overexpression


Case Study 1

Breastcancer.org

CA
15.3


ER/PR+


Digital
Tomosynthesis
 Papillary


Colloid


Thermography


CA125


Subtrac(on
Probe
Technology
 Mucinous
 Comedo
 Progesterone
Receptors





Cribriform




Chromogenic
In
Situ
Hybridiza(on
 Ki‐67


Her2/neu


Oncogene
Overexpression
 CA
27.29


Case Study 1

Breastcancer.org ER/PR+


Digital
Tomosynthesis
 Papillary


Colloid


Posi(ve
margins


CA
15.3


Thermography


CA125


Ductal
Carcinoma
In
Situ


Ductal
Lavage
 Subtrac(on
Probe
Technology
 Mucinous
 Comedo
 Progesterone
Receptors


Cribriform







Chromogenic
In
Situ
Hybridiza(on
 Ki‐67


Her2/neu


ImmunoHistoChemistry


Oncogene
Overexpression
 CA
27.29
 Fluorescence
In
Situ
Hybridiza(on


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…

Case Study 1

Breastcancer.org

Case Study 1

Breastcancer.org Lessons from User Research… Clear navigation and powerful search were not enough But, oh, the risks of personalization…

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

Case Study 1

Breastcancer.org

Five vocabularies were developed to assign metadata to content.

Audience


Situa(on


Clinical
 Characteris(cs


Perspec(ve


Topic


Case Study 1

Breastcancer.org Audience

  Patients   Family & Friends   Press & Public   Clinicians & Providers   Worried Well

Case Study 1

Breastcancer.org Situation

  Just Diagnosed   Waiting for Test Results   Undergoing Treatment   Recovery & Renewal   Metastatic Cancer   End of Life

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

Case Study 1

Breastcancer.org Perspective

  Clinical   Emotional   Practical   Press / Public

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

Case Study 1

Breastcancer.org Audience


Patients

Situa,on


Undergoing Treatment

Clinical
 Characteris,cs


Lymph Nodes Removed: 20+

Case Study 1

Breastcancer.org Audience


Patients

Situa,on


Undergoing Treatment

Clinical
 Characteris,cs


Estrogen Receptor+ Progesterone Receptor+ Pre-menopausal

Case Study 1

Breastcancer.org Audience


Patients

Situa,on


Undergoing Treatment

Audience
 Clinical
 Characteris,cs


Estrogen Receptor+ Progesterone Receptor+ Pre-menopausal

Case Study 1

Breastcancer.org Metadata for personal profiles was made possible by adding structure to already in-use signature information

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

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.

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.

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.

Case Study 2

Data Warehouse Client The “Person”  

Pharmaceutical corporation sales staff

Case Study 2

Data Warehouse Client The “alization”  

Custom ColdFusion application

 

PostgreSQL database

 

iText Java Library PDF rendering tool

Case Study 2

Data Warehouse Client How it works Sales reps start by selecting:   Drug Class  

Drug

 

State

 

Health Plan

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

Case Study 2

Data Warehouse Client How it works  

Sales reps may add “personal touches”, and then generate a printready PDF file.

 

Automatically sent to approved print vendors for production.

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.

Case Study 3

ShipYourReptiles.com

(Snakes on a Plane)

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

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

Case Study 3

ShipYourReptiles.com The “alization”  

Custom Ruby on Rails application

 

PostgreSQL database

 

Google Analytics

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

Case Study 3

ShipYourReptiles.com How it works
 Account managers may review order activity, and upgrade accounts for highvolume customers.

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.

Conclusion: Review of Key Steps of Personalization:

Measure! Develop and Test Automate

Questions?

Questions? Derek Olson VP, Foraker Design [email protected] (303) 449-0202