Upgrade to Pro
— share decks privately, control downloads, hide ads and more …
Speaker Deck
Features
Speaker Deck
PRO
Sign in
Sign up for free
Search
Search
An Introduction to A/B Testing
Search
timminor
October 03, 2011
2
520
An Introduction to A/B Testing
Barcamp talk given at UX Camp Brighton 2011.
timminor
October 03, 2011
Tweet
Share
More Decks by timminor
See All by timminor
UX Quiz, UX Camp Brighton 2014
timminor
0
130
A UX Quiz, v2
timminor
0
160
A UX Quiz
timminor
1
510
Thinking fast, thinking slow: designing for Cognitive ease
timminor
5
1.7k
Featured
See All Featured
Docker and Python
trallard
45
3.5k
How To Stay Up To Date on Web Technology
chriscoyier
790
250k
Into the Great Unknown - MozCon
thekraken
40
2k
A Tale of Four Properties
chriscoyier
160
23k
Mobile First: as difficult as doing things right
swwweet
223
9.9k
10 Git Anti Patterns You Should be Aware of
lemiorhan
PRO
656
61k
Reflections from 52 weeks, 52 projects
jeffersonlam
351
21k
Statistics for Hackers
jakevdp
799
220k
Become a Pro
speakerdeck
PRO
29
5.5k
Easily Structure & Communicate Ideas using Wireframe
afnizarnur
194
16k
Learning to Love Humans: Emotional Interface Design
aarron
273
40k
For a Future-Friendly Web
brad_frost
179
9.9k
Transcript
@"mminor An Introduction to A/B Testing Photo credit: @lucyjspence
and h7p://www.flickr.com/photos/futureshape/4530140521/
What we’ll be looking at • What is A/B Tes"ng?
• What to test? • What is A/B Tes"ng good for? • A/B tes"ng should not replace • Do’s and Don’ts • Tools
Source: h7p://www.smashingmagazine.com/2010/06/24/the-‐ul"mate-‐guide-‐to-‐a-‐b-‐tes"ng/ What is A/B Testing?
What to test? • Bu7ons: colours, sizes, loca"ons •
Headlines • Product descrip"ons • Form fields (adding and removing) • PPC landing pages • Email content • Pricing and promo"ons • Images
What is A/B Testing good for? • Data-‐driven design
• Ge\ng stakeholders on-‐board • Avoiding personal preferences
A/B testing should not replace: • Talking to users
• Usability tests • Acceptance tests
Do’s and Don’ts • Test simultaneously • Test o^en
• Include only new visitors • Show repeat visitors the same version • Use results to inform future development • Require sta"s"cal significance DO:
• Test too many ideas at once • No
clear idea • Don’t let your preconcep"ons override results • Peek at your results and stop as soon as a significant result has been achieved. See: Evan Miller. Do’s and Don’ts DON’T:
The tools Other tools are available: A/Bingo, Vanity, Unbounce,
Performable, Vertster, SiteSpect, Webtrends Op"mize and Omniture’s Test&Target
The Quiz Show
The Corkscrew A
The Corkscrew B
The Corkscrew A/B
A
Great American Days A
Great American Days B
Great American Days A/B
B
Real Gap
Real Gap A/B
B
CityCliq A
CityCliq B
CityCliq A/B
A
Vendio A
Vendio B
Vendio B
A
Quick Fire Round • What were the two main tools
men"oned? • What percentage increase did the winning RealGap test achieve? • Name one of the three things A/B tes"ng doesn’t replace.
What we looked at • What is A/B Tes"ng?
• What to test? • What is A/B Tes"ng good for? • A/B tes"ng should not replace • Do’s and Don’ts • Tools
References • First slide photograph: h7p://www.flickr.com/photos/futureshape/4530140521/ • www.google.com/websiteop"mizer
• h7p://visualwebsiteop"mizer.com/ • h7p://www.abtests.com/ • h7p://whichtestwon.com/ • h7p://www.smashingmagazine.com/2010/06/24/the-‐ul"mate-‐guide-‐to-‐a-‐b-‐tes"ng/ • h7p://www.conversion-‐rate-‐experts.com/ar"cles/101-‐google-‐website-‐op"mizer-‐"ps/ • h7p://www.evanmiller.org/how-‐not-‐to-‐run-‐an-‐ab-‐test.html • h7p://elem.com/~b"lly/effec"ve-‐ab-‐tes"ng/ • h7p://remoteresear.ch/