So you want to grow your eCommerce sales that means You need more traffic right?

You need to get more links, spend more on Adwords and post more frequently to social media – don’t you?

Traditionally this is how businesses have grown their sales online, however this approach is running into some serious issues today. Google is more unpredictable than ever, Adwords are getting expensive and your Facebook Reach is at an all time low.

So if more traffic is out of the question how do you grow your sales?

The answer – Conversion Rate Optimization.

A B Split Testing

What is Conversion Rate Optimization?

Conversion Rate Optimization (or CRO for short), is a systematic process to increasing the percentage of website visitors that convert into customers or leads.

For eCommerce businesses like yourself the goal of CRO is to increase your overall sales, sale value and lifetime customer value.

Commonly CRO is done through a series of ‘split tests’ or ‘A/B tests’. These are tests where you adjust 1 portion of a web page while keeping everything else equal. Your website traffic is then split 50/50 between the 2 pages and results are monitored by specialised software

The result is a statistically significant result that either indicates that the test helped, or didn’t help.

Once you’ve got this data you can decide if you’d like to implement the changes on your website as a permanent fixture.


Why every eCommerce business needs to invest in CRO

Unlike traditional marketing where more traffic meant more expenses, CRO is about helping you extract more value from each of your current subscribers. This results in a higher visitor value, and ultimately boosts your overall profitability.

To better explain how CRO can help your bottom line, I’ve included an example below:


How CRO Works


In this example the overall conversion rate of the website has increased from 2% to 3%. Now while this mightn’t sound like a large increase, it is a 50% boost and as a result will increase in 50% more sales from the same volume of traffic.

Some business owners will leave it here and take a healthier profit, while others will then increase the amount that they spend to buy traffic and increase market share. Ultimately this depends on your situation and how you’ve structured your business model.


The 6 elements of  Conversions

During a recent webinar that we through with Hunter Boyle from Optimization Co-Pilot, he explained the 6 elements of CRO extremely well which you can watch in the webinar excerpt below:


CRO Distilled

Originally distilled by the guys over at Marketingexperiments, the 6 elements that affect conversions are broken into 3 sections:


Elements that increase conversions

  • Relevance – The more relevant your message is to what your audience is seeking, the higher your conversions will be. For example if you have an advertisement about blue shoes, and they land on a page filled with multi-coloured shoes then your conversions will drop. If however the landing page is all about blue shoes & has them all listed for sale, then you’ll see your conversions rise.
  • Clarity – Making your offer & website clear and easy to use will boost your conversions. Cluttered sites, unclear shipping or return policies will all negatively affect your conversions

Elements that decrease conversions

  • Anxiety – Your aim is to minimize client anxiety as much as possible. Do this by making your site easy to use, and by helping them find what they are after as easily as possible.
  • Distraction – When someone is in your sales funnel, you want it to be like a greased slide. You want it to be completely friction-less and free of distractions. This means if a message isn’t relevant, remove it. For example most checkout pages still have menu bars, where this adds distractions that will lower your conversions.

Elements that accelerate conversions

  • Clear Value Proposition – Chances are there are dozens of websites selling the same thing as you, so what makes you stand out? By having a very clear value proposition on all crucial pages of your website you’ll accelerate the sale and help them take action sooner.
  • Urgency – As humans we are lazy by nature. That means unless we need to do something, then chances are we’ll put it off. By utilizing urgency as a part of your CRO strategy you can put a serious fire under your prospects and boost conversions dramatically. Kogan do a wonderful job of this with both Urgency and Scarcity in play throughout the site. They also work social proof in extremely well.


The 4 steps to building a CRO strategy


Now that you understand what CRO is and how useful it can be for your business, let’s get you upto speed with how you can get started today.

Before you rush into any split test and start making changes to your website, there are 4 steps that you must follow to build a powerful, effective CRO strategy

Measure your website’s performance

In order to improve your websites conversion rates, it’s essential that you know where you are starting. You want to know what is happening and why it’s happening.

This is where Google Analytics comes in.

If your Google analytics is properly set up then you’ll have plenty of data to work with to measure your website’s performance. If however it isn’t then get it set up effectively ASAP so you have several weeks to collect your data.


What is your business objectives?

The easiest way to discover this is to ask yourself the question ‘ Why does my website exist’? As this post is all about eCommerce CRO I’m going to assume that your website’s goal is to make sales.


What are your website goals?

Goals are strategic in nature and are derived from your business objectives. For example as an eCommerce business  you may have the following goals:

  1. Do A – Add better product descriptions
  2. Increase B – Increase click through rates
  3. Reduce C – Reduce abandon shopping cart rate

These goals are your priorities which are expressed as simply as possible. its essential that you are clear on your outcomes before we start diving into your data. Its also essential that these goals are properly set up in Google analytics so we can track how you’re going with them.


Define your KPIs

In order to effectively track your progress we need clear, concise numbers to hit. To do this ensure that you set your Key Performance Indicators as a numerical metric.

For eCommerce stores KPI’s are usually…

  • Number of products sold
  • Total sale volume
  • Average order volume
  • Net Conversion  Rate
  • Lifetime Customer value


Define your target metrics

Say your store sold 83 items last month. Is that good? Or absolutely terrible? For your KPI’s to mean something you need to set target metrics.

Start by defining a target for every KPI that is important to you. For example:

  • Number of products sold = 200/mo
  • Total sale volume = $20,000/mo
  • Average order volume $100
  • Net Conversion  Rate = 2.5%
  • Lifetime Customer value = $500


Get Qualitative data

Digging into Google Analytics can only get us so far into understanding your customers mind set. Its great at telling us what happens but it doesn’t shed much light onto why it happens.

For this its best to speak to your customers directly and to get their feedback. Now I’m not saying to call every customer but rather to get creative with surveys and focus groups.

Some examples to help you collect more qualitative data include…

  • Adding an exit survey to your site asking visitors why they didn’t make the purchase with you
  • Perform usability testing with a focus group to get real time feedback
  • Send a feedback survey to your clients both after sale as well as if they don’t purchase to find out more about their motives.
  • Add an exit survey to the thank you page asking those who did buy why they made that decision.


2. Prioritize your testing opportunities


Once you have your metrics and your goals in place, the next step is to prioritize what you test. This prevents you from spending hours and hours optimizing your ‘about us’ page only to later realise that very few people visit it. It also helps you to determine the most lucrative opportunities to start with.

By digging into your Google Analytics data such as ‘top landing pages’ and ‘user flow’ you’ll get to see where your traffic is, as well as where people are dropping off.


Prioritize pages with high potential

One of the easiest places to start is pages that have high traffic, but low conversions. Sometimes this will be your home page, but most likely it will be a deep category page. I recently was doing an analysis for a client and found that 2 specific categories had drop off rates 3 times that of others – This is a prime example of a page to prioritize.


Top exit pages

These are the last page that someone see’s before they leave your site. By ‘plugging the leaks’ so to speak you can recover a lot of sales. Commonly eCommerce businesses have very high drop off rates around the cart and checkout process – so this is a great place to start.


Prioritize tests based on value and cost

In the ideal world resources wouldn’t be an issue, but the reality is that you only have so many hours in a day and so much budget to invest in testing, so its critical that you get the best ROI possible.

So before you set any tests in motion, prioritise your tests based on the value and cost.

Widerfunnel has a great framework called PIE that helps with this process:

    1. Potential – How much potential does this page have to increase your conversion rate?
    2. Importance – Is this page important? Does it have much traffic? Will it affect customers? Is the traffic quality any good?
    3. Ease – How easy is it to implement the test? Are there technical or political barriers around testing this page?

For each of these sections give them a rating out of 10. Then average the results to get your PIE number. Then simply start with the highest PIE test and work your way through the list.

To help you understand fully below I’ve included an example table:

PIE Measurement Example

Framework Table image from Widerfunnel

3. Testing

The third section of the CRO system is your testing phase. This is where the magic happens and the rubber hits the road. This is where all of the in depth analysis that you’ve been doing comes together and goes from being simple data into actionable hypotheses and tests.


Form a clear hypothesis for your test

Hypothesis for eCommerce CRO

A effective Hypothesis will determine the success of your overall CRO results.

A Hypothesis sums up your test in a few simple sentences. It states the problem as well as how you’re going to fix it.

They need to be clear, concise and easy to understand. This is because once your test is under way and you start questioning what you’re trying to achieve, it will keep you on track.

A good hypothesis needs to be…


    1. Testable – You need to have a specific metric that you can test
    2. Have a goal of solving a problem – Split testing is done to solve your specific problems
    3. Gain market insights –  Other than increasing your conversion rates & profits, split testing will also give you data about your clients. A well thought through hypothesis will allow your split test results to give you in depth information about your customers that will be helpful in the future.

The importance of testing to statistical relevance

If you flip a coin 10 times in a row, there is a pretty high chance that you’ll get 7 heads or 70% heads. Where if you flip the coin 1,000 times in a row, you’ll likely get 490-510 heads – or 49-51%. So why is the percentage so much different?

The answer is statistical significance. You see, when you test with a small amount of traffic too many things are left up to chance. This is why when running your split tests it’s essential that you get at least 1,000 hits to that particular page.

I also suggest that you aim for a statistical significance of 95% or above before you decide on a winner. This means that your results are based on maths and hard numbers not chance.

Most good software such as Optimizely and VWO will calculate this for you.


Test for profit & revenue

At the end of the day you’re in business to make a profit. There are a lot of sites out there that will tell you that testing for revenue is the way to go, and that as long as your sales revenues grow then everything is hunky-dory.

However if you’re boosting your revenue by discounting your items, then your profits and margins are shrinking and your bottom line could be worse off than before you started testing.

This is why I always suggest carefully implementing your test and ensuring that you have your analytics set up professionally to guarantee that your getting proper data to make your decisions on.


What to test – the low hanging fruit

When testing your site it’s always best to come up with your own hypotheses to get the best results, however sometimes a checlist is helpful. So below are a few examples of what to test if your stuck:

  • Call to action buttons – Size, colour, wording and placement
  • Forms – Length, field times, text on forms
  • Copywriting – headlines, guarantees, value propositions, product descriptions
  • Layout – home page, product pages, landing pages
  • Checkout – Menu, call to action, steps, instructions

What software to use

When implementing tests I highly recommend that you use a professional split testing software such as Optimizely or VWO.

They are both great software’s, however I personally prefer Optimizely. It gives you more control over your testing environment and structure, plus they have a basic account that is free.

Once you get experienced I highly suggest upgrading to their Enterprise accounts (which aren’t that much in comparison to your ROI) and work with an experienced company to get the best results.


4. Learn from your results and start over


David Ogilvy, possibly one of the greatest marketers of all time said over 70 years ago “Never stop testing, and your advertising will never stop improving”. This is the #1 rule that all business owners and marketers should follow daily.

Just because you’ve finished 1 split test, that by no way means that your site is at its full potential. Heck – If you run 5 tests a month for a year you still won’t be at full potential. To be successful long term in eCommerce you need to test on an on-going basis to ensure that your profitability is as high as possible. Because ultimately whoever is most profitable can run the other players out of a market.

This is why once you’ve finished your first test, revisit your PIE analysis and start planning your next test.

Not every test will grow your conversion rates, however if you look hard enough there is a valuable lesson in the results of every test that will help you in the future.


A/B Testing Considerations

SEO considerations

Google openly endorses split testing, because its goal as a business is to ensure its users find what they are looking for. So the better you make your website, the happier their customers will be. However there are some SEO implications for split testing that you should be aware of.


    1. Use Rel=’canonical’ – Instead of using a noindex tag or adjusting your robots file add a canonical link to your website. This will tell search engines that this is a variant of your original page and will prevent them from indexing both versions.
    2. Use 302 NOT 301 – NEVER 301 redirect a page unless you are sure that you want to permanently delete the original. To simply send visitors to a new variation during your test run a 302 redirect as this is Google’s ‘temporary’ version.
    3. Don’t run your tests longer than you need to. – Once you have a winning version swop running your tests. This will prevent any unexpected penalties from occurring.
    4. Use a proper software to prevent penalties from cloaking- When you are running a test through a proper software such as Optimizely, Google can recognise what is going on and will be more lenient on its anti-cloaking rules.

Forget about A/A testing

Occasionally you’ll hear people speak about A/A testing to ‘calibrate’ their software. If you’re using a professional solution there is no need for this and it simply wastes your precious resources.


Avoid the percentage confusion

When quoting your results speak in terms of ‘change as the amount by which one variation is larger than the other’ rather than ‘ change as the numerical difference between the two’.

For example, if you increase your conversion rate from 2% to 3% you boosted your conversion rate by 50% – not my 1%.


Mobile considerations

With Mobile being a highly important aspect of eCommerce these days, it’s essential that you compare your tests in mobile views as well.

Deal with your web design agency while implementing tests and they’ll ensure that your tests fit within the RWD framework of your site.



As you can see, Conversion Rate Optimisation can be a very lucrative investment for any eCommerce business owner It can grow sale exponentially and position your business so that you can truly scale with ease.

However in order to implement a successful CRO campaign management takes careful planning and knowledge. Hopefully this guide has armed you with everything that you need to know to get your first test online in the next week.

If however you’re still a little confused, or would like a hand then please feel free to get in touch and we’d be happy to help you out.