Intelligent InterREaction: An argument for a datadriven approach to UI design

1
Intelligent Inter(RE)actionAn argument for a
data-driven approach to UI design

• Garrick Schmitt, Vice President, User Experience
• Marisa Gallagher, Director, User Experience San
  Francisco
• Avenue A I Razorfish

2

• 1. Data Must Become Part of
• Our Design Vocabulary

Early Web Offline Methods Evolution of a
Data-Rich Platform
3

• 2. We, as Digital Designers, Need to Evolve

New Tools New Methodologies New Ways To
Measure Success
4

• 3. The Future of Design
• Will Be Data-Driven

Analysts Are Royalty Fail to Succeed Rise of
Intelligent Systems
5
It is not the strongest of the species that
survive, nor the most intelligent, but the ones
most responsive to change -- Charles Darwin
6

• The Early Days of the Web Were All About Bringing
  Things That Worked Offline, Online

7

• NYTimes
• Circa 1996
• Pages
• Links (22!)
• Forums
• TOC

http//www.archive.org
8

• NYTimes
• Circa 2007
• Pages
• Links
• Search
• Images
• Multimedia
• AJAX
• RSS
• Personalization

http//www.nytimes.com
9

• Yahoo!
• Circa 1996
• Pages
• Hyper Links
• Search

http//www.archive.org
10

• Yahoo!
• Circa 2007
• Pages
• Hyper Links
• Search
• Images
• Multimedia
• AJAX
• Web 2.0-ness

http//www.yahoo.com
11

• Offline Design Methods Worked

Observe
A-Ha!
Design Launch
12

• Ideo
• Circa 1999
• Process
• Observation
• Brainstorming
• Prototyping
• Results
• Brilliant Ideas
• Elegant Design
• Awards
• Fame

13
Shoppers could use the cart like home base,
darting down an aisle with a basket. At checkout,
clerks could pack the groceries in plastic bags
that neatly hook within the frame. As far as we
know, no one had done anything quite like it
before. To me that's the heart of it, a real
innovation that redesigns the shopping
experience. -- Tom Kelley, The Art of Innovation,
Ideo
14
So here's a question what happened to the IDEO
shopping cart? Is anyone using it? After six
years, my local grocery stores are still using
wire basket carts. Superstore, the groceropolis
around here, has carts that fit plastic
baskets--but they have more in common with
ordinary carts than the IDEO cart. -- Gene Smith,
Atomiq
15
Digital Data
Today we watch videos on YouTube, chat with
friends through a host of I connect with people
on MySpace a and post our photos to the
world on Flickr. . All of this is all
rendered dynamically on top of one of the most
impressive data-rich platforms ever imagined.
16

• So Are We Prepared to Embrace Data?
• Do We Know What to Do With It?

17

• How Do We, as Digital Designers, Need to Evolve?

18

• Data-Driven Design

• New Tools
• New Methodologies
• New Success Metrics

19
Primary Research Data-Mine Customer Touchpoints
20

• Usability Testing Clickstreams Click-Density

• Click Density
• Link Analysis
• Field Analysis
• Scrolling Analysis
• Within-Page Drop-off Analysis
• Between Page Drop-off Analysis
• Flash Tracking
• Integration of attitudinal surveys with
  behavioral data

21

• Product Release Rough, Right Ready For Change

In Praise of Multivariate Testing!
22

• Success Metrics Quantifiable With Business
  Impact

Conversion Rates
Phase 1 Changes 21
Phase 2 Changes 28!
Step 1 Arrived On Page
Step 2 Interacted With Page
Step 3 Started Form
Step 4 Submitted Form
Step 5 Converted
23
The traditional Microsoft-approach is like
state socialism, with the Mandarins from the
best schools locked in an ivory tower on a nice
campus setting coming up with five-year plans,
formulating big APIs and then having it built
from scratch -- Adam Bosworth, Google
From Intelligent Reaction, Salesforce.com
AppExchange Conference http//www.salesforce.breez
ecentral.com/intelligentreaction/
24

• Those Who Understand the Data Will Design the
  Future

25

• Fail Faster to Succeed More!

26

• Its a Beta-World Rapid Releases and Ongoing
  Optimization are here to Stay

MoMB
The Museum of Modern Betas
http//momb.socio-kybernetics.net/
27
Why
Intelligent Inter(RE)action?
28

• Intelligent Systems will React to Real User
  Behavior in Real-Time

29
The Adaptive User Interface Machine Learning
30

• 5 Ways To Prepare For the Future
• Validate User Personae With Real User Data
• Demand Quantitative Online Testing (Not Just
  Usability)
• Define Quantitative Success Metrics Early
• Accept That There Are No Right Answers, Just
  Ongoing Testing Refinements
• Become a Data-Detective Find new inputs,
  sources and testing mechanisms

31
Thanks )

• Garrick Schmitt (garrick.schmitt_at_avenuea-razorfish
  .com)
• Marisa Gallagher (marisa.gallagher_at_avenuea-razorfi
  sh.com)