Digital Growth Day: September 18, 2014
CONVERSION RATE OPTIMISATION FROM THE GROUND UP
So, you've built a website that attracts a fair few visitors each month. Conversion rates are okay, but you know there’s room for improvement. Actually knowing how to make improvements, however, is a very different beast.
This session looks at how to turn your thoughts of running a CRO campaign, into building an action plan that will help:
-Gather feedback from real customers and learn what they think about your website
-Come up with your very first testing ideas
-Ensure your results are being tracked accurately
-Avoid some of the common pitfalls of running CRO campaigns
Tim Allen, Consultant, Distilled
8. @tallen1985
We spend on:
SEO
Paid Advertising (Google, Bing, or Social
channels)
Email Marketing
plus many more channels
9. What if you spend all this
money but your conversion
rate sucks?
@tallen1985
https://medium.com/@kitinmaki/does-your-conversion-
rate-suck-c9b1077f6990
10. @tallen1985
If you’re selling nuclear
submarines you probably won’t
be selling that many
Craig Bradford
Senior Consultant, Distilled
23. You should look to run tests for at least two business
cycles and at least 250 conversions (ideally 350+)
@tallen1985
http://www.slideshare.net/sullivac/20-simple-ways-to-fuck-
up-your-ab-testing-measurefest
24. How long is it going to take
to get this many
conversions?
@tallen1985
40. @tallen1985
GOOGLE CONSUMER SURVEYS
WHATS THE ONE THING THAT WOULD STOP YOU
BUYING CURRENCY ONLINE?
Answers:
“Security”
“Is Paypal accepted”
“Knowing my details are safe”
“Being sure my money will arrive safely”
http://www.googleconsumersurveys.com
55. Know the scale of your
changes - this will influence
the way you create your
tests
@tallen1985
http://moz.com/blog/6-things-i-wish-i-knew-before-
using-optimizely
69. @tallen1985
Statistical significance is all
about whether the difference
between two numbers is
meaningful or just a fluke
Tom Capper
https://www.distilled.net/resources/statistical-significance-
for-cro-6-things-you-need-to-know/
77. DISCOVERY
- SEGMENT DATA TO LOOK FOR CAN’T CONVERT
OPPORTUNITIES
- GATHER AS MUCH USER DATA AS POSSIBLE TO
UNDERSTAND WHY CUSTOMER WON’T CONVERT
78. EXPERIMENTS
- ENSURE ANALYTICS TRACKING IS TURNED ON
- DON’T CHANGE TRAFFIC ALLOCATION ONCE TESTS ARE
LIVE
79. RESULTS
- ONLY ROLL OUT SUCCESSFUL TESTS THAT ACHIEVE
SIGNIFICANCE OF 95% OR MORE
- USE TRAFFIC ALLOCATION TO ROLL OUT VARIATIONS
WHILE WAITING FOR DEVELOPERS TO PERMANENTLY
IMPLEMENT CHANGES