Conversion Rate Optimisation: The Science Behind Turning Visitors Into Customers

James Spittal (Me)

• Built first HTML website at age seven
• Some of the companies I’ve worked
with:
– Ning, Nando’s, Pearson Australia,
Deloitte, Thomas Duryea Consulting,
ANZ, Melbourne Water, Master Builders
Association of Victoria, State Library of
Victoria
– Lots of small businesses

• I’ve spoken at CEO Institute, CPA
Australia, Swinburne University, Hub
Melbourne, SEO events and a few other
places
• I’m the CEO & founder of digital agency
Web Marketing ROI
• Consult/advise on digital for a number
of different businesses
• I also own and operate a number of
different eCommerce websites and
online web properties

• Most of what I’ll be talking about today is stuff
I’ve directly implemented for clients of mine
or things that I’ve used in my own business.
• There’s a lot of ‘theory’ out there about
conversion optimisation .. And sometimes I
find there is a bit of a gap between the theory
and the actual practical application.

How this talk is relevant to you &
outcomes
• What I’d like you all to walk away with is a clear
understanding of how you can, by applying Conversion
Rate Optimisation:
–
–
–
–
–

(A) Generate more leads online; and/or
(B) Generate more sales online;
Drive (or increase) business growth by using (A) and/or (B);
Apply this to your client’s businesses;
Apply this to your own business;

• Can’t teach everything in one presentation and this is
the tip of the iceberg, but I’m hoping to cover as much
as I can about #CRO!

What Actually Is Conversion Rate
Optimisation?

Is it a report about how you can
improve your conversion?

I know. Let’s Google the answer.
• Wikipedia:
The method of creating an experience for a website or
landing page visitor with the goal of increasing the
percentage of visitors that convert into customers. It is
also commonly referred to as ‘CRO’.
• Unbounce:
Conversion rate optimization (CRO) is a discipline that
focuses on the testing of landing pages, aiming to find
the optimal design and messaging mixture that
produces higher conversion rates.

Source: http://www.jamesspittal.com/blog/conversion-rate-optimization-definition-what-exactly-is-conversion-rate-optimization/

So why will some website redesigns
ruin our conversion rate and others
will improve it?

Which landing page converts better?

Full credit to whichtestwon.com for this example

Version A increased submissions by
24% (read: 24% more leads) by
removing the image from the page.
(no, that doesn’t mean we should
always remove images)

Which converts better?
Full credit to whichtestwon.com for this example

Version A (without the video!)
increased leads by 64%
(no, that doesn’t mean you should
always not use videos.)

Common A/B Testing Assumptions
• ‘Red converts better than any other colour!’
• ‘Having a video always converts better than not
having a video!’
• ‘Having a headline is always better than no
headline!’
• ‘Long copy beats short copy!’
• ‘Short copy beats long copy!’
• ‘The best design wins!’
• ‘Less form fields is ALWAYS better than more
form fields!’

Take-Away
• Unless you’re testing – you are just guessing.
• Ignore the myths and hype about Conversion
Optimisation. Test and don’t fall into the error of
making tests.
• You can guess – but make sure you back it up with:
– 1). More than hundreds of A/B tests and thousands of
whichtestwon.com style tests
– 2). An actual A/B test. What works for one industry may
not have the same effect on another industry.

A/B Testing
• ‘So if we just do heaps of A/B testing,
we’ll always be right and always
make our clients lots of money and
conversion improvements, right?’

Is Your Client Even Ready For A/B
Testing?

https://github.com/jamesspittal/ab-testing-calculator

PHP open source tool I created for A/B testing statistical significance
calculations. I also created it to illustrate the danger of small sample
sizes.

How well do you know your statistical significance
mathematics? Do you expect this will be more than 99.95%
statistically significant or not?

Nope, it’s not! And it also violates the <1000 visitors rule.
Therefore, ending this test would provide scientifically
uncertain conclusions and false negatives! Don’t do it.

Typical Example
•
•
•
•

Monthly Google AdWords Spend $4,000
Google AdWords Cost-Per-Click $4
$4000 / $4 = 1000 clicks per month
‘Typical’ conversion rate of clicks ->
conversions: 6.8%
• 1000 clicks = 68 conversions
• Sale price in this example: $100
• 68 x $100 = $6800

How could we change the original
website (‘landing page’) to increase
the conversion rate?
(The number of people signing up on
that page.)

A/B Testing
Google AdWords Visitors

50% visitors to the
‘original site’

50% visitors to the ‘new
test site’ (optimised)

In a scenario such as this, it’s realistic to expect a 15% - 150% increase in conversion rate
on Version B over Version A.

Before & After
• Monthly Google AdWords
Spend $4,000
• Google AdWords Cost-PerClick $4
• $4000 / $4 = 1000 clicks per
month
• ‘Typical’ conversion rate of
clicks -> conversions: 6.8%
• 1000 clicks = 68 conversions
• Sale price in this example:
$100
• 68 x $100 = $6800

• Monthly Google AdWords
Spend $4,000
• Google AdWords Cost-PerClick $4
• $4000 / $4 = 1000 clicks per
month
• Increased conversion rate:
13.6%
• 1000 clicks = 136
conversions
• Sale price in this example:
$100
• 136 x $100 = $13,600

Net result:
100% increase in sales revenue
generated via Google AdWords
without spending an extra cent on
advertising or marketing.
Best of all, in most cases – A/B
testing can be setup in a couple of
days.

Real examples
• Over a pool of approximately 500 A/B tests and
approximately 100 clients, I’ve seen increases in
conversion rate of 15% - 600%
• I’ve never seen an example where there wasn’t
opportunity for at least 15% improvement in
conversion rate, and the average over 100 clients
is around the 150% mark
• Average length of engagement was
approximately 12 weeks – but those that had the
highest increases tended to be closer to 12
months (diminishing returns)

Lessons learned
• The tests that you think are going to increase
conversion often don’t! (Consumers don’t always
behave the way we’d expect or the way we’d hope.)
• The most important thing is to test.
• If the test produces a negative result, come up with a
hypothesises why you think it did and repeat another
test.
• Continue until you see a positive result.

The Process: Continuous & NeverEnding Testing Loop

An example of a successful A/B test
(100% uplift)

Variation #1 (Winner)

100% increase in leads without increasing advertising expenditure

• Visual Website Optimizer
screenshots, etc.

Where a lot of companies fall over in
implementing these strategies

Where a lot of companies fall over in
implementing these strategies
• Who’s responsible for
this stuff again? (Is it the
marketing manager? Is it
the digital marketing
manager? Is it the IT guy?
Is it the website guy? Is it
the agency?)
• Do we have the ‘know
how’?

• Are we getting bogged
down with other things?

Source: Conversion Rate Optimization Report 2012

“The 12 Commandments Of #CRO”
•
•
•
•
•
•
•
•
•
•
•
•

1. Never make assumptions.
2. Never guess.
3. Never assume that what worked for one industry will work for another.
4. Always be improving your #CRO skills by constantly learning from tests
and improving your % of guesses being accurate.
5. Regularly measure your own accuracy on whichtestwon.com.
6. Don’t bother testing for clients who don’t have enough conversions.
7. Calculate the estimated testing duration before beginning.
8. Follow the conversion principles – not the tactics.
9. Repeat after me: 99.95% statistical significance.
10. The bigger the conversion improvement, the shorter the test duration.
11. Align the A/B testing programme directly with client expectations in
terms of improvement, past tests run and time lines.
12. Conversion Improvement != Equivalent Revenue Improvement

If you execute this correctly,
improvements of 200% - 300% in
less than 3 months is possible.
Yes. That means that:
- Cost-per-lead would be reduced by up to 300%
- Cost-per-sale would be reduced by up to 300%
- Sales revenue from lead generation would be increased by up to
300%
- Online sales (E-Commerce) would be increased by up to 300%

Q&A
• Questions?
•

[For the online readers, feel free to e-mail at james@jamesspittal.com and I’d be happy
to answer any questions. I actually answer every e-mail I get (worst case, you might
have to wait a week or two if I’m really busy.)

•

For convenience, put the subject as: “SlideShare Presentation: Conversion Rate
Optimisation”]

Thank you!

Me
• Tweet me: @jamesspittal

• Growth hacking/CRO
consulting gigs with me
personally:
http://www.jamesspittal.com

• CRO (and web/SEO/PPC)
engagements and RFPs:
http://www.webmarketingroi.com.au

Thank you!

Web Marketing ROI
•

We work with clients in the US, UK, Australia
& Hong Kong

•

Full-service Digital Agency that specializes in
Conversion Rate Optimisation

•

We also offer all the usual suspects: SEO,
AdWords, PPC.

•

We have designers, programmers,
copywriters, wire framers, A/B testers.
Everything you’d expect from a full service
digital agency.

•

Existing relationships with many agencies
around the world (A/B Testing, Reseller,
Affiliate, Etc.)

•

I’m the Founder & CEO and without Web
Marketing ROI, I wouldn’t have the spare
time or resources to play with this stuff!

Conversion Rate Optimisation: The Science Behind Turning Visitors Into Customers

More Related Content

What's hot (20)

Similar to Conversion Rate Optimisation: The Science Behind Turning Visitors Into Customers (20)

Recently uploaded (20)

Conversion Rate Optimisation: The Science Behind Turning Visitors Into Customers

Editor's Notes