Making the Business Case for Usability

1
Making the Business Casefor Usability

• Kathryn Summers
• 2003

2
Fundamental UCD activities you want someone to
pay for

• An early focus on users and tasks
• Empirical measurement of product usability (ease
  of learning, ease of use)
• Iterative design

3
Usability goalsthe golden ring

• Usefulness (enables user to achieve her goals)
• Effectiveness, or ease of use (can sometimes be
  defined quantitatively)
• Learnability
• Likeability

4
Common Objections

• We already know what our users want.
• Were changing the process (or this is a new
  process), so the old process is irrelevant.
• Were just changing one part -- study that.
• All users are different. We cant design it to
  please just a few users.

5
Standard business case

• Description of project
• Market analysis
• Costs and benefits of business as usual vs.
  business as proposed
• Personnel and equipment costs
• Project timetable
• Analysis of project dependencies
• Risk analysis (reasons and probability that costs
  could be higher, benefits lower than projected)

6
Measuring ROI increased benefits

• More sales
• Higher productivity
• More successful proposals
• More awards
• More tasks completed
• More people take medicine as prescribed
• Higher customer satisfaction ratings

7
Measuring ROIreduced costs

• Fewer support calls
• Fewer complaints (perhaps even less litigation)
• Reduced need for training
• Reduced maintenance/repair costs
• Less downtime for workers
• Less effort (time, lines of code, rework)
• Fewer errors, accidents

8
Product owner vs. user costs

• Costs to site/product owner
• increased support costs
• reduced sales
• expensive redesign
• Costs to customer/user
• reduced employee productivity
• increased employee frustration
• more frequent mistakes

9
Cost of support

• Hypothetical case
• Software application costs 150
• Support call costs 9 a minute (very reasonable
  assumption)
• So a single support call could mean you lose
  money on that sale

10
Effect of usability on sales

• Humans like to feel competent. They dont like to
  feel incompetent.
• High error rates make users uncomfortable, and
  can lead to avoidance of the application.
• Usable products get recommended to other users.

11
Employee turnover

• Employee turnover is expensive. Estimates are
  that the cost of turnover is 1.5 times the
  employees annual salary.
• Improving employee job satisfaction and reducing
  turnover can dramatically increase profits.
• Tech support employees
• Customers and other users

12
Shortening development cycles

• User research can help you identify which
  features matter most to users
• You identify key tasks more reliably and earlier
  in the development process.
• Adding features and fixing errors costs ¼ as much
  in the early stages of development as it costs in
  later stages. Post-release changes cost even more.

13
Research

• Schlesinger and Heskett, 1991.
• Total cost of employee turnover is 1.5 times
  annual salary.
• Mantei and Teory, 1988.
• Early changes cost ¼ as much as late changes.
• Wixon and Jones, 1992.
• 40 of development effort is user interface.
• Pressman, 1992. Martin and McClure, 1983.
• 80 of software lifecycle costs are post-release
  maintenance. 20 of these costs are to fix bugs
  or reliability problems, 80 are to support unmet
  or unforeseen user requirements.

14
User research techniques

• User interviews
• User questionnaires
• Usage statistics
• Contextual interviews
• Task walk-throughs task scenarios and personas
• Card sorting
• Image elicitation or image collaging
• Qualitative and quantitative usability testing

15
Calculating costs for user research and usability
testing

• Equipment and materials
• Travel expenses
• Participant compensation
• Employee work hours
• Cost of employee work hours must be the loaded
  cost hourly rate of worker plus benefits plus
  office space plus equipment plus support
  personnel costs

16
Your costs for contextual interviews

• Planning interviews
• Recruiting participants
• Conducting contextual interviews
• Your time, travel
• Participants time
• Analyzing results
• Preparing report

17
Usability testing costs

• Writing scripts and screeners
• Recruiting users
• Conducting pilot test
• Conducting tests
• Your time, travel
• Participants time
• Video equipment and materials
• Analyzing results
• Preparing report
• Preparing video highlights

18
Survey costs

• Developing survey
• Conducting pilot test
• Conducting survey
• Paper, mailing
• Web, hosting, database
• Analyzing results
• Reporting results

19
Prototype costs

• Designing prototype
• Building prototype
• Testing prototype (see usability testing)
• Revising prototype
• Re-testing prototype
• Creating style guide

20
Limitations of usability testing

• Testing situations are always artificial, and
  thus will influence the results
• Test participants are usually not perfectly
  representative users
• The test design can never fully duplicate natural
  user behavior

21
Common questions

• How much more usable will it be?
• How much more usable does it need to be?
• How do we measure the usability?
• How much will it cost to make it more usable?
• How much more money will I make (or save) if we
  spend money on usability?

22
Calculating BenefitsReduced Processing Time

• Data Entry
• 250 users
• 60 screens per day
• 230 work days per yr
• 1 second reduction per screen
• 25 loaded hourly wage

• Benefit per year
• 250
• x 60
• x 230
• x 1/36000
• x 25
• ______
• 23,958 per year

23
Calculating BenefitsReduced training

• Data entry training (one week reduced by four
  hours 10)
• 250 users
• 4 hours less
• 25

• One-time benefit
• 250
• x 4
• x 25
• _______
• 25,000

24
Calculating BenefitsDecreased Errors

• Data Entryone fewer error per week
• 250 users
• 0.2 errors eliminated per user per day
• 2 minutes recovery time per error
• 230 work days per yr
• 25 loaded hourly wage

• Benefit per year
• 250
• x 0.2
• x 230
• x 0.833 per error
• ______
• 9,580 per year

25
Calculating BenefitsFewer late design changes

• Changes made early cost ¼ as much as late changes
• 20 early changes
• 8 hrs per change
• 45 loaded hourly rate
• Early changes
• 20x8x457200

• One-time benefit
• Late changes
• 4x720028,800
• 28,800
• -7200
• 21,600

26
Total savings

• For first year
• 23,958 (reduced processing time)
• 25,000 (reduced training)
• 9,580 (reduced errors)
• 21,600 (fewer late design changes)
• Total 80,138
• For subsequent years
• 23,958 (reduced processing time
• 9,580 (reduced errors)
• Total 33,538

27
Conservative Estimates

• Make very conservative estimates because they are
  hard to challenge
• Measure before and after benchmarks to show that
  benefits were much greater than estimated
• Next proposal will be welcomed with open arms

28
Cultural barriers

• Management
• One department pays costs of user research while
  another department reaps benefits
• Development team
• Want to do work that is fun
• Want to do work that is easy
• Want to re-use old code
• Want to code in modules
• Want to make the system run more smoothly and
  efficiently
• Think they are just like users, but arent
• Customers/buyers are not users

29
Triggers for change

• A high-level advocate
• A big, expensive usability pratfall
• Market pressures
• Education
• Internal, small-scale, proven ROI

30
Further resources

• Marianne Rudisill, Clayton Lewis, Peter B.
  Polson, Timothy D. McKay, eds. Human-Computer
  Interface Design Success Stories, Emerging
  Methods, and Real-World Context. San Francisco
  Morgan Kaufman, 1996.
• Randolph G. Bias, Deborah Mayhew, eds.
  Cost-Justifying Usability. New York Harcourt
  Brace, 1994.
• Janice Redish, Judith Ramey. Special section
  Measuring the value added by professional
  technical communicators. Technical Communication
  42.1 (Feb 1995) 23-83.

31
Conclusion

• User research and usability testing do cost
  money.
• Choose what types of research to do, and when to
  do it, based on what will give you the best
  return on your investment.