A year of awakening? Despite the strong business case for making website improvements that maximise revenue, core challenges preventing many companies taking CRO to the next level.
Creator Influencer Strategy Master Class - Corinne Rose Guirgis
The state of conversion rate optimisation 2019 [updated]
1. The State of Conversion Rate
Optimisation (CRO) 2019:
a year of awakening
2. The business case for CRO
Effects of higher conversion
rates include:
o Increased customer acquisition,
revenue & net profit
o Increased ROI on current
marketing activities
Set the foundation for future
marketing endeavours:
o Google Adwords
o Facebook Ads
o Etc.
The uplift from
optimized web pages is:
o Immediate
o Ongoing
Studies show that improved
user experiences:
o Increase retention and
repeat business
o Increase lifetime customer
value and reduce churn
4. Compound effect
1 + 1 = 3
CRO has a compound effect.
Consider the following scenario:
Homepage + 19% uplift
Category page + 19% uplift
Product page + 19% uplift
Shopping cart + 19% uplift
If you achieved these results, you would have
doubled your conversion rate, because each
improvement compounds upon the previous one.
5. What can CRO do?
Acquisition
Maximise your…
Sales
Revenue
Average order value
Revenue per visitor
Money metric...
6. What can CRO do?
Retention
Customer lifetime value
Money metric...
Repeat purchase rate
Maximise your…
8. Ecommerce companies: mature at CRO?
63% lack a structured optimisation process*
It’s critical to have a framework, otherwise you’re winging it
74% of CRO’ers say that a structured approach to conversion improved their sales.*
* Source: Econsultancy 2018 Optimisation Report
9. Hypothesis generation
83% of companies run experiments without
supportive data*
Because we saw [data/feedback]
We expect that [change] will cause
[impact]
o A hypothesis without evidence support is just a
guess!
o CRO is a marriage of qual/quant data findings
that inform hypotheses:
We’ll measure this using [data metric]
*Source: Econsultancy 2018 Optimisation Report
10. Evangelisation and Culture
Changing the status quo is hard
Common barriers include project management,
interdepartmental silos, and executive buy-in
o Siloed organization (cited by 38% of practitioners)*
o Conflict of interest between different
departments (cited by 31% of practitioners)*
Navigating office politics can be the difference
between incremental growth and explosive growth.
*Source: Econsultancy 2018 Optimisation Report
Two key barriers to adopting a culture of experimentation
11. Statistics: is it really understood?
There is no research that interrogates the relative understanding
of statistics and its application to A/B testing
o Was there sufficient sample size? How long did the test run for?
o Was statistical power considered?
o Was the calculation of minimal detectable effect evidence-based?
o What was the primary success metric? Was it binomial or
non-binomial?
12. Statistics
You can’t do CRO without statistics
False positives:
you think the A/B test is
positive but it’s not
Because you run the risk of:
False negatives:
you think the A/B test is
negative, but it’s not
False positives are more detrimental.
False negatives are more common.
Testing without applying statistics
is like wearing a blindfold
14. Focus on test velocity
The success of your testing program is a combination of three factors:
Those three combine to indicate test velocity.
o Number of tests run (quantity)
o Percentage of tests that provide a win (quality)
Combine with sample size and impact per
successful experiment, and you get an idea of total
business impact.
o Sequence of tests (priority)
15. Breaking down (A/B) test velocity
Remember, your test velocity comprises of the
number of tests run (testing volume), how many
tests will win (quality of your hypotheses), and
test sequence (priority)
Test velocity holds you accountable
TEST VELOCITY =
Number of tests run x
% successful tests x
sequence
16. Test Velocity: sum of
many parts
A healthy test velocity is indicative of...
o Support of senior leadership
o Culture of testing and experimentation
o Structured optimisation process
o Evidence-based hypotheses
o Application of statistics (positive A/B tests have
a financial impact)
17. Conclusion
Near-ubiquitous CRO adoption is not matched by competency
Hard skills: statistics, user research (empathy)
o Everyone has the data and the tools
o Limitations are people-based:
Soft skills: aligning teams, changing culture
o Resource, budget and executive buy-in remain the biggest constraints
o These is what differentiates world-class optimisation programmes
from the rest
See my article: https://hostdgtl.com/blog/mind-the-skills-gap/
18. Want to maximise
website revenue?
Get in touch:
o https://linkedin.com/in/chapmanalan/
o hostdgtl.com
o I’m Alan Chapman, website conversion expert, speaker on the art of selling and resident
course trainer at Digital Kitchen, the digital marketing trainer provider.
o If you want results-based consultancy that maximises website returns, I’m here to help.
o FREE ebook: dive deeper into the concept of 'Velocity’, the 3-point formula that increases
visitor-to-customer conversion rates, marketing ROI and online revenues.
○ Get the 38-page ebook here
20. Mythbuster #1: Be Sceptical of Best Practice
CRO results are seldom uniform across websites and verticals
o Results from one site don’t always apply to another
o Instead of making assumptions, test different page variations and rely on data to make decisions
o Again, it comes back to data research
See my article: https://hostdgtl.com/blog/rise-of-conversion-patterns/
21. Mythbuster #2: CRO ≠ A/B testing
CRO a multidisciplinary approach to persuading visitors to action
o There is a common misconception that split-testing is the final output of CRO
o The final output is a change to your website
o Sometimes that change is validated with a A/B test that confirms that the change is a positive one. However, sometimes the
change is rolled out without A/B testing
See my article: https://hostdgtl.com/blog/case-against-website-ab-testing/
o The decision whether or not to A/B test should be taken on a case by case basis