Designing for the customer and the bottom line

Designing for the customer and the bottom line: Costs and benefits of user-centred design Gitte Lindgaard Carleton University Ottawa, Canada

• To convey your …. message you must know what that message is and how to convey it (Chapanis, 1988) 2

Traditional UCD focus • User experience – Who are the application users? – What tasks will the application support? – What other tools do users employ? – What is the environment like? 3

UCD cost-benefit analysis • 10 years ago • Estimated costs of applying usability methods (time to set up, run tests, analyse data, write up) • Estimated benefits of finding x number of usability problems 4

UCD cost-benefit analysis now • What are you selling? • Who are you selling to? • Can you substantiate your claims? • Hypothetical numbers are low in credibility • Case studies may not look like they are similar to the current situation • Is the time right? 5

Who is our immediate client? • Marketing: promote snazzy features • But don’t give them ideas that IT can’t deliver • IT: technical wizardry and miracles • But must be linked to the business case • Project manager: deliver on time and to budget • Finance: keep costs within budget • Training: cannot increase training costs • So must know the implications of your great idea • Corporate ID: must reflect our image 6

Who owns the user experience? • • 7 Hardware. . . Shape, size, weight, labels… Software … installation procedure Ok? QA …fault free products Tech writers … user manuals, instructions Training … sufficient, economic CSRs … cope with customer questions Marketing … position products etc.

8 Customer services Training Software Tech writers QA Hardware Different owners & cultures

What’s your business? • Consumer products – Generic hardware/software (PDAs, games, smart watches) – For a specific target audience (business DSS, financial planning SW, scheduling SW) • Customer-specific products – Internal to your organization – For outside customers (EMS, WMS) 9

What’s your business? • Consumer products – Generic hardware/software (PDAs, games, smart watches) – For a specific target audience (business DSS, financial planning SW, scheduling SW) • Customer-specific products – Internal to your organization – For outside customers (EMS, WMS) 10

Targeted consumer products • Dental support system • • 7 manyears > $2 million development costs Worked with dentists … employed at a Unversity Special features: – – 11 Multimedia data storage & manipulation Video camera … to confirm diagnosis Speech I/O … to avoid keyboard Excellent scheduling capability … if > 4 dentists

User instructions • Problems installing a printer driver • • 50, 000 calls to help desk + send out patch on disk Cost > $900, 000 + between $600, 000 and $12, 5 mill Evidently no user testing prior to release • User manual style and template • 50% of $1, 68 mill per year • Cost of style & template $80 K 12

The Cordis story 1996 1995 1993 • $223 m Annual sales • $20 share value • <1% market share of angioplasty balloons • Goal: 5% market share 13 • $443 m annual sales • Stent was born. . . • $1 billion sales in year 1 • Johnson & Johnson acquired Cordis at • $109 share value

14

Know thy users’ goals • Cordis focused on desired outcomes • • Outcome-based interviews Desired outcomes Decompose the process Organise the outcomes in categories corresponding to each step in the process • Rate outcomes for importance and satisfaction • Use outcomes to jump-start innovation (Ulwick, A. W. , 2002) 15

Outcome-based interviews • Began to define every step underlying usage process • • 16 Inserting catheter into an artery Placing the balloon at the lesion or blockage Opening the artery by inflating the balloon Removing the catheter

Stakeholders, yes but… • Know precisely who to focus on • Cardiologists who perform the procedure • Nurses who assist the procedure • Hospital administrators who sit on the purse strings 17

Capture desired outcomes • Moderator • • clarifies and validates customers’ statements digs beneath the surface of words makes sure every aspect of the process/activity is gone through Each statement translates into one desired outcome containing – Type of improvement required (increase minimize) – Unit of measure (time, n, frequency) – Outcome statement used later in benchmarking, competitive analysis, concept evaluation – Rephrases statements until they are free of solutions – Then confirms statements with participants 18

Organise the outcomes • List collected outcomes, remove duplicates • Categorise into groups corresponding to each step in the process, here • Inserting catheter into an artery • Placing the balloon at the lesion or blockage • Opening the artery by inflating the balloon • Removing the catheter • With this we can measure value to the customer 19

Rate and use outcomes • Rate outcomes for • Importance and satisfaction, i. e. how well is each outcome currently satisfied? • Use outcomes to jump-start innovation • Cordis identified several new product opportunities and set new goals (e. g. 20% reduction in restenosis) 20

[Importance + (Importance – Satisfaction) = opportunity] Desired outcome Importance Satisfaction Opportunity Minimize restenosis 3. 2 15. 8 Minimize force to cross lesion w. balloon 8. 3 4. 2 12. 4 Minimize damage (dissection) 9. 5 7. 5 11. 5 Minimise time to place balloon 9. 1 8. 4 9. 8 Minimise time to complete procedure 21 9. 5 5. 1 1. 0 9. 2

Changing UCD focus • Who are the application stakeholders? – Includes users • What are their goals? – Includes tasks • What products compete with this one? – Includes other tools used 22

Product life cycle Amount of effort Time 23 Product 1 Product 2

Now, to the web Fortune 1000 companies spend on average $1, 5 - 2, 1 mill/year redesigning their web sites without knowing whether this actually improves them for users Kalin (1999) 24

Overall goals 1 a Business goals • Customer conversion • Increase retention • Increase transactions 2 1 b Customer goals • Easy to learn • Easy to find & buy • Access to help when I need it 4 User interface 3 Back-end system and processes from Donahue (2002) 25 Satisfaction of Customer goals leads To satisfaction of Business goals Engagement and Interaction process, Including content and transactions

Novice Target user zone New user -Tentative -May need proactive help from Donahue (2002) 26 Early adopter - fast - aggressive - needs little /no help - values fast, efficient user experience

Web development life cycle Architecture Information and Technology architecture Visual design Look & feel 2 3 4 Validation Strategic planning usability tests 1 Define user and business goals 5 Execution 8 Business evolution 6 7 Maintenance from Donahue (2002) 27 Launch

Deliver value to all segments Business goals User Convert browsers to customers and Increase client base Prospective customer (i. e. browser) Online experience must drive Relationship establishing activity Directed user behaviours Researching firm requesting information and follow-up Increase Retention rate Existing customer Relationshipretaining behaviour Increasing frequency of engagement, level of membership Deepen existing Customer relationships Existing customer Relationshipdeepening activity Increase number of points of contact with customer Creation of usable data about profitable and cost-efficient transactions Action based on understanding and analysis of transaction and data usage Increase productivity Employee Inside the firm inside the firm from Donahue (2002) 28

What’s your business? • Consumer products – Generic hardware/software (PDAs, games, smart watches) – For a specific target audience (business DSS, financial planning SW, scheduling SW) • Customer-specific products – Internal to your organization – For outside customers (EMS, WMS) 29

Hidden costs • In the U. S. 31% large development projects (budget > $100, 000) cancelled before completion • Of those that were completed, 52. 7% were 189% over budget on average and cost an additional $59 billion in 1995 (Ewusi-Mensah, 1997) • A 6 -month delay in release adds 33% aftertax profits (House & Price, 1991) 30

Creeping costs • Fixing a problem worth $1 during project definition costs • $10 if fixed during development, and • $100+ if fixed after product release 31

Creeping costs Cost of changes N possible design alternatives Requirement Phase 32 Development Phase Deployment Phase

Creeping costs • Total maintenance costs $20 -30 billion/year • Backlog maintenance minimum 167% of this (Martin & Mc. Clune, 1998) • … 80% of all problems are identified during maintenance, and it is estimated that 80% of these could be saved with a thorough UNA! 33

Calculating costs and benefits • 2 parameters are important: – Cost/benefit ratio • Relationship between projected use of resources and post-implementation/post-sale value of the investment (and savings during development) – Payback period • The amount of time it will take before the product starts to outweigh the cost of the investment 34

Costs and benefits – Cost/benefit ratio • Relationship between projected use of resources and post-implementation/post-sale value of the investment (and savings during development) – Cost • Team of 4, 2 hours @ $60 = $480 – Saving • 4 person days = 32 hours @ $60/hour = $1, 920 – Cost/benefit ratio = 1: 4 35

An example • Core system - customer records – customer requests – services & equipment for every telephone number (n lines, n sockets, cables, exchanges) – linked with other systems (billing, cables, faults, WP/YP, service orders, scheduling) 36

OEO 50 I ANDREW MEDICI I 11489 _____ EE#: _____ INIT ENTRY: SERV#: _______ EXT#: *** TPIS/2 ORDER S&E ENTRY *** EXCH ACT: J EXCH CODE: ______ INPLACE: _____ IP DATE: ______ EXCH RMK: __________________________________ SERVICE NAME: ________________________________ ADDR INFO: ________________________________ SUB ADDR TYPE: _______________ SUB ADDR#: __________ STREET#SFX: ________ NAME: __________ TP/SFX: _______ LOCALITY: _____________ POSTCODE: ____ FLT CLR PTY: ___ TEMPORARY: _ AL CONN: _______ ROT CONN: ___ PEXT? __ LINE TYPE: _____ NEXT FAC: _ LN SEL: __ ARE CODE: REM/CCN(R/N): __ - SERV STATUS: ___ DISC TYPE: PAL#: _______ RENTAL RATE: ___ METER RATE: ___ CONC AUTH: _____ CAB#: ______ ASIC: _____ DIRECTORY ACTION REQ (Y/N): _ DIRECTORIES#=-> WH: __ YE: _ COM: ___ COMMAND: ______ 37

OEO 50 I ANDREW MEDICI I 11489 _____ EE#: _____ INIT ENTRY: SERV#: _______ EXT#: *** TPIS/2 ORDER S&E ENTRY *** EXCH ACT: J EXCH CODE: PBX 334 INPLACE: 3374606 IP DATE: 04089996 EXCH RMK: N/A QREMM SERVICE NAME: PETER JAMES HOUSTON ADDR INFO: ROWANVILLE MELA SUB ADDR TYPE: SEMI-ATT. DUAL H/H DWELLING SUB ADDR#: 337 A STREET#SFX: CRESCENT NAME: ST BARTHOLOMEW TP/SFX: PMT/YY LOCALITY: ROWANVILLE WEST POSTCODE: 3957 FLT CLR PTY: ROP TEMPORARY: N AL CONN: P/T ROT CONN: 3 F PEXT? Y LINE TYPE: PBX NEXT FAC: N LN SEL: PR ARE CODE: TTYP-M REM/CCN(R/N): 2 X - SERV STATUS: INP DISC TYPE: PAL#: POTT/P RENTAL RATE: RES METER RATE: PM 1 CONC AUTH: MELA CAB#: 228 FS/476 PW ASIC: 338 PY DIRECTORY ACTION REQ (Y/N): N DIRECTORIES#=-> WH: N YE: N COM: PB COMMAND: EXT 38

Through the operator’s eyes • Activity analysis of 2, 000+ calls (Wilson, 1995) – Typical call activities: Interacting with computer Interacting with customer Walkabouts Making phone calls Paperwork Other behaviour Total time wasted 39 190 sec 148 sec 33 sec 28 sec 24 sec 12 sec 33 min/d 28 min/d 24 min/d 85 min/d

Other problems • Data accuracy rate (for accepted transactions only) = 70% • Database not in place at cutover • On-the-spot system rejections = ? % • Absenteeism increased >200% during study – 2 suicide attempts, 2 nervous breakdowns, numerous staff on long term stress-leave • Staff turnover rates ~120% in 3 months 40

CSR time $M 13 Cost Communication $M 1. 6 Tech time $M 3. 7 CSR Training $M 10. 4 Help, support $M 0. 5 Total $M 30!!! 41 Enhancements $M. 8

• Cost-benefit ratio: – Total costs of waste year 1 $ 30 million – Total cost of study $ 90 K At least 50% could be fixed easily at a cost of $1 million and within 1 year – Waste after 1 year – Cost-benefit ratio 42 $15 million 1: 1. 38

Another example • HR processing applicants – – Current process time 4 hours/applicant Data entry costs $25/applicant $100/applicant HR processes 1, 000/year $100, 000 Programming time 40 hrs @ $60/hr $ 2, 400 • Usability work shows 50% user time saved, but say 25% – – – – 43 Cost/applicant now $75 or $75, 000/year Benefit per applicant $25 or $25, 000/year Benefit in year 1($25, 000 -$2, 400) $22, 600 Benefit in year 2 $25, 000 Benefit in year 3 $25, 000 Total benefit (year 1 + 2 + 3) $72, 600 Cost-benefit ratio 72, 600: 2, 400 1: 30. 25

What’s your business? • Consumer products – Generic hardware/software (PDAs, games, smart watches) – For a specific target audience (business DSS, financial planning SW, scheduling SW) • Customer-specific products – Internal to your organization – For outside customers (EMS, WMS) 44

Warehouse Management System • Bottlenecks: Assemblers • • Handheld devices sometimes miss an entry Voice Recognition System freezes …Assemblers work to standards …. they can’t offload orders in bay specified • Bottlenecks: Loaders • • 45 Can’t find orders Loads are too tall for trucks Loads must be dismantled and reassembled … 2/3 of loaders’ job spent re-doing assemblers’ work

46

Changing UCD focus • Who are the application stakeholders? – Includes users • What are their goals? – Includes tasks • What products compete with this one? – Includes other tools used • What is the context like? – Includes community of work, division of labour, rules 47

UCD opportunities • Strategic planning (web) • Product definition stage (Cordis) • User Requirements capture stage (dental support system) • Outside of development cycle (Customer records system) • During SDLC (traditional usability studies) 48

UCD methods • Social/behavioural science methods – Opinion-based (interviews, surveys, reviews, focus groups) – Performance based (activity analyses, user tests, observations) 49

UCD challenges • Must expand – User focus to business focus – Single product to product range focus – Understanding of business goals, processes – Notion of “team” beyond project team • Must move outside the usability box 50

Lindgaard’s Backyard November 2002 51

Lindgaard’ s Backyard May 2003 52