Why you're always wrong: User Research for Start Ups

User Research for startups@ bitspiration 2013, Krakow

@bendressler

PsychologyMSc Maastricht University

Games

Usability Analyst Shop Direct Group

User Research

United KingdomGermany

SwedenThe Netherlands

UK‘

sbiggesthomeshoppingcompany

very.co.ukLittlewoods.comisme.comKandco.comfashion

electricalsjewellery

furnituretoys

beauty

billionrevenue

£ 1.7 $ 2.6 € 2.0 of which80% is generatedonline

40%of allsessions

mobile

UX

Tempting ways to design

by committee

by inspiration

by user request

A better way to design

by observing

what your users

can easily use

User research…

…is a methodology to figure out how wrong your assumptions are

You are smarter

because…

You are not aware of that

Your observations are biased

What they expect

What they wantHow they want it

don‘tWhat they can‘t

What they knowdon‘t

don‘t

What they perceivedon‘

t

How wrong are you?

How severe is that?

amount of wrong

Layer 1

LanguageIntelligence

MemoryPerception

Motor skills

ContrastsColors

Movement HearingSpatial

Verbal

Procedural

Logic

Theory of mind

Processing speed

RecognitionRecall

Self awareness

Reflexes

Aiming Coordination

Deciphering

Producing Interpretating

Layer 2

This is a link

Every browser has a „back“ button

Grayed out means „inactive“

Tabs are independent browser windows

The browser sits on the computer and displays websites

X usually means „close this“T:\this\is\a\path

Mouse cursor changes indicate potential actions

Layer 3

The contacts sit in the header

Rows full of blocks are deleted

All confirm buttons are orange

You have to enable location services

The circle button makes the avatar jump

Swipe with all fingers is different

Logout sits under settings

You can turn off push notifications

The cake is a lie

How big is your problem?

How unique is the proposition?

How tech-savy

are users?Reddit

eBay

Farmville

League of Legends

Skype

Photoshop

iPhonefacebook

SPSSPoker tracker

World of Warcraft

Techcrunch.com

9gag.com

Google

Linux Pebble Watch

Ouya

Virality

Conversion

MAUNPS Bounce

Open wallet conversion

RetentionCheckout dropout

Average order value

Friends invited

Average lifetime value

Not sure how?

Qualitative QuantitativeData level

Resourcesrequired

Clicktracking

Live survey

Biometrics

Controlled experiments

Eyetracking

Card sort Web analytics

Survey

Clicktest

Diary study

Techcrunch.com

User testing

Google Key task analysis

Split testing

Session replay

Cognitive walkthrough

Qualitative QuantitativeData level

Resourcesrequired

Clicktracking

Live survey

Biometrics

Controlled experiments

Eyetracking

Card sort Web analytics

Survey

Clicktest

Diary study

Techcrunch.com

User testing

Google Key task analysis

Split testing

Session replay

Cognitive walkthrough

Qualitative QuantitativeData level

Resourcesrequired

Clicktracking

Live survey

Biometrics

Controlled experiments

Eyetracking

Card sort Web analytics

Survey

Clicktest

Diary study

Techcrunch.com

User testingGoogle

Key task analysis

Split testing

Session replay

Cognitive walkthrough

„About that research budget...“

Top 5 tasks

Supports prioritisation

Representativeness of users crucial

Key Task Analysis

No cost

User tasks required

Difficult to simulate lack of knowledge or skill

Cognitive Walkthrough

All about observing

Possibility to prompt or ask

Most major issues with five users

User Testing

Check for false affordances

Pen & paper solutions

Quantify results

Clicktest

Open for ideation

Closed for validation

Gather verbal feedback

Card Sorting

Analytics required

Failing fast

Does not answer „why“

Bigger samples needed

Split Testing

How to make magic happen

Triangulate!

Iterate!

BadBetterGoodWorse

Start early!

Be creative!

Challenge accepted!

1. You are wrong

2. Test your assumptions early

3. Choose the right method

4. Triangulate

5. Iterate

@bendressler