Sorting the Cards in Customers Favor: Improving the Information Architecture and the Tool

1
Sorting the Cards in Customers Favor Improving
the Information Architecture and the Tool

• Kerrie Green, Hoovers Online Usability Manager
• Sharon Harap, TNS Intersearch Consultant

UPA 2002 Conference Poster Revolution July 10,
2002
2
Background

• Business information provider Hoovers Online
  (www.hoovers.com) needed to simplify its
  information architecture.
• Hoovers wanted to use feedback from customers
  and prospective customers as part of the redesign
  process.

3
The Problem

• Specifically, Hoovers wanted to know which site
  content categories were meaningful to users and
  what content could be grouped logically within
  those categories.
• The site decided to run a card sort and wanted to
  include national participants, so it needed to
  perform the project online.
• Timeline
• 5 weeks
• But there was not enough in-house programming
  time or usability bandwidth.
• Hoovers also wanted an easy-to-use interface
• No free tools were terribly simplistic.
• Set-up on free tools was bear-ish.
• Hoovers IT staff didnt have time to program the
  free tools and couldnt provide support for them.

4
Card Sort Folders Categories

• Site Categories (Folders)
• Company Overview
• Industry Research
• In-Depth Company Research
• People Research
• Not Important

• Selected Site Content (Drag-able
  Items)
• Key Products
• Business Description
• Business Units
• Contact Information
• Sales/Financials
• Key Customers
• Breaking News
• Board Members
• Downloads
• Suppliers/Vendors
• Current Trends
• Organizational Structure
• Strategy Research
• Ownership

5
Review of Available Tools

• Hoovers had one week to review pre-existing card
  sort tools and to set up a remote sort.
• Some of the tools were free or low-cost but
  provided little or no technical support. They
  also werent flexible, easy to use, or easy to
  set up.
• Cost-benefit analyses on such tools as IBMs
  EZSort package and WebCAT, developed by the
  National Institute of Standards and Technology
  (NIST), showed the free tools were known to be
  buggy and had clunky interfaces. They also
  werent intuitive to set up and were not visually
  appealing to users.

6
The Solution

• Hoovers contracted to build a custom remote card
  sort and survey, which enabled the project to
  move swiftly and meet the companys research
  objectives.
• Contracting to customize the tools was just as
  efficient and cost-effective as devoting internal
  staff to develop and launch the free tools.
• Other companies could also
• Invest 11,000 to hire a research firm with
  specific knowledge of card sorts and online
  application design and development. Fees include
  consultation, programming, field, analysis, and
  reporting.
•   Design a custom online card sort built
  specifically using Macromedia Flash.
• Provide feedback on the user interface, including
  design, layout, and the placement of terms,
  folders, and instructions for use.

7
Project Implementation

• Timeline
• Internal project, 5 weeks
• Provider contract, 3 weeks
• Card sort launched and closed, 1 week
• Results returned, 1 week after close
• Study Design
• Hoovers usability staff helped design and
  customize the new tool.
• Tool was built in Macromedia Flash.
• Card sort involved a simple design of click-able
  terms to be dragged into folders.
• Design retained Hoovers branding so the
  experience would feel familiar to participants.
• Online screener questions targeted specific
  non-customer groups for the study while customers
  were handpicked and invited to participate by
  e-mail according to their target profiles.
• Participant feedback was gathered during some
  portions of the sort.

8
continued

• Survey Recruitment
• Hoovers target customers received e-mail
  invitations to participate in the study, which
  they could then launch from the invitation
  itself.
• Prospects could opt to participate in the study
  while using the site. Once screened, they could
  proceed to the card sort and complete the study.

9
Interactive Card Sort Tool
Multiple topics could be dragged up to three
times.
Users could drag topics into folders.
10
Key Findings

• Developing a custom online card sort provides a
  great deal of flexibility.
•   Having a branded study was valuable for
  achieving a high response rate.
•   Linking to the Web site at the end of the study
  offered users a seamless experience.
• Consider providing an option to have respondents
  label the categories or provide alternative
  naming options.
• Creating a custom card sort can be as
  cost-effective as, and even more efficient than,
  producing one in-house.

11
Summary of Analysis
12
Comparative Analysis of Online Tools

• IBMs EZSort Software Package
• EZSort uses statistical cluster analyses to
  organize data. Participants use its USort program
  for remote studies, while administrators use the
  tool to create studies and enter results. They
  also use the EZCalc program to manage data and
  perform analyses.
• Free.
• Confusing set-up.
• Requires individual test data entry.
• Inefficient and time-consuming.
• Huge learning curve.
• Not visually appealing.
• Not fully customizable.

• NISTs WebCAT
• Performs remote card sorts to determine if users
  understand site categorizations. The card sorts,
  implemented as Java applets, are delivered online
  or by e-mail.
• Buggy.
• New features category-naming option and text
  area for user comments.
• Difficult set-up.
• Not visually appealing or fully customizable.

13
Interactive Card Sort (Developed by Hoovers
Research Provider)

• Customizable and flexible features
• Implemented in Macromedia Flash.
• Retains product look-and-feel and branding
• Familiar for customers.
• Can be delivered from host site or by e-mail.
• Developed with companys business objectives in
  mind.
•  Inexpensive -- Price varies, but about 11,000
  per implementation with survey.
• Easy to Use
• Participant instructions are clear and easy to
  follow.

14
Summary of Interactive Card Sort Benefits

• Fast delivery.
• Inexpensive.
• Remote.
• Customizable and visually appealing.
• Simple interface and instructions.
• Ability to deliver surveys and collect user
  feedback.
• Quick results.
• Analysis performed by usability and market
  research professionals.

15
Study Tips

• Educate research provider with detailed
  background on project-related usability and
  business objectives.
• Leave labels open on categories or offer labeling
  options.
• Make word tags large enough to grab and drag into
  categories.
• Eliminate scrolling.
• Make sure participants fit target customer
  profiles.
• Make sure study objectives support overall
  business goals.

16
Poster Presentation from UPA 2002
17
Contact Information

• Kerrie Green
• Hoovers Usability Manager
• Austin, Texas
• Phone Numbers
• Work 512-374-4683
• Cell 512-293-6413

• Sharon Harap
• TNS Intersearch Consultant
• Austin, Texas
• Phone Number
• Work 512-346-3075