A CRO strategy is a repeatable set of actions to make sure more of your existing traffic carries out the behaviour you want them to. It focuses on analysing data (quantitative and qualitative), applying user psychology to ideate and formulate hypotheses, prioritising, testing hypotheses and repeating (it also includes a rollout of changes if a test was successful). This can be roughly summarised as a test and learn process.
There are different ways to approach a conversion rate optimisation strategy – this is ours
There are lots of conversion rate optimisation experts out there and lots of opinions about what a CRO strategy looks like. This is the CRO strategy we follow at Credo, and I hope you will find value in learning about it.
Before you dive in, define your goals
The first step in your CRO strategy is to agree on what goals you are trying to achieve. There is more to conversion rate optimisation than just improving conversion rates. For example, you may want to:
- Drive down your cost of customer acquisition.
- Increase your revenue.
- Increase uptake of a free trial.
- Increase general user engagement.
It’s important that everyone is aligned on what the goals are as these help with hypotheses formulation (ideas for testing). It’s also important to understand that not every conversion goal has to be big.
You may have what we call ‘micro-conversions’, little steps that bring the user closer to the overall conversion. Uptake of a free trial is likely one of these micro-conversions. Others include things like:
- Downloads of a whitepaper
- Subscribing to a newsletter
- Signing up to a webinar
- Adding to basket.
You should be able to allocate macro (large) and micro-conversion goals to each step in your user journey and to each page on your site.
Step 1 in your CRO strategy – analysis
Traditional data sources
It is easy to think analysis = Google Analytics, and you’re not wrong, it is a great way you can analyse the behaviour of users on your site. If you need some tips on where to start with Google Analytics, check out this article. It’s important to look at different types of users in Google Analytics, such as those who came from one traffic source compared to another, and new users versus returning users.
Your analytics data may not be stored in Google Analytics, it may be stored in other tools like PostHog or Fullstory (two platforms I’ve been using for clients recently). You may have sales data you want to analyse in Shopify, or another data source such as existing test data where you can see what worked well and what went wrong.
Heatmaps and recordings
Alongside these typical quantitative (numerical) data sources you can review quantitative data in heat maps like click maps and scroll maps. Tools like Hotjar (click here for 30 days free), and Microsoft Clarity are great at helping you uncover the numbers that traditional data sources often won’t show you without a bespoke setup. Things like what percentage of users are clicking on a CTA versus how many are scrolling to that CTA, and what percentage of users are reaching your social proof are great applications.
Watching recordings of what your users do within these pieces of software is also an option. Although slightly time-consuming, it can be a great form of qualitative data.
User research and desk research
In an ideal world, you will have actually spoken to your users before coming up with ideas for testing via user research. This is because as clued up as you think you are, nothing comes close to hearing the truth from your target audience’s mouth.
Heuristic analysis of your own site and competitor analysis
Perhaps the least scientific, (but none-the-less valuable if done right) is a heuristic evaluation of your own site and of competitors’ sites. It is best to have an external party do this for you to remove the veil of internal knowledge in the form of a CRO audit.
This way you get a much clearer view of what someone coming in fresh sees, and to understand why issues are problematic on a deeper psychological level (as well as some suggestions on how to fix them).
Step 2 in your CRO strategy – ideation and prioritisation
If you’ve had a CRO audit done, much of the ideation part may be done for you. This is where suggestions are made based on the available data and user psychology. If not, it’s time to get creative and problem solve the issues you’ve uncovered.
There are a few different prioritisation frameworks out there. The MoSCoW framework for prioritisation stands for Must-have, Should-have, Could-have, and Won’t-have (or similar). It requires the key stakeholders including the product team coming to a consensus on what’s important.
Personally, I like the ICE scoring method too, which stands for impact, confidence, and ease. How I like to use it is to give each ideation a score for impact, confidence, and ease and then multiply these together. This allows me to see which ideas are good candidates for including in or constituting a hypothesis. I put these all into he backlog of a project management Trello board.
Step 3 in your CRO strategy – hypothesis formulation
Hypothesis formulation is a little different to ideation as it lays out your theories in such a way that tie to your goals. ‘If X happens, Y will happen’. I like to also add ‘as a result of Z’, to make it really clear why I believe the hypothesis will come true.
The hypotheses you choose to work on first should be those connected to high priority ideas for testing. You may need to combine a couple of ideas to make a hypothesis work, but you shouldn’t combine too many as this makes it difficult to see what worked and what didn’t once you have run a test.
For example, you may decide to change the header on your landing page at the same time as adding bullet points summarising your selling points because you believe that focusing all messaging above the fold (the bit people see when a page loads) on the value you offer will decrease bounce rates.
You won’t know if it was the messaging change or the bullet points that made the difference, but you might be OK with this as the two were aligned with the same rationale (the ‘Z’ from before). You may also feel you need to do this if you have limited traffic and testing one thing at a time would simply take too long. Once you have your hypotheses you may decide you want to apply a prioritisation framework to these too.
Step 4 in your CRO strategy – testing
Testing is probably the thing most people think of when they think of conversion rate optimisation. Unfortunately, testing is often rushed into, ignoring all the previous steps which laid the groundwork for tests that are likely to succeed.
When you are ready to test it’s important you pick a good piece of testing software. We recommend our partners Shoplift for testing on Shopify sites and Webtrends Optimize for testing elsewhere. The support from both is great if you are new to testing.
When you think you are ready to start a test, check that:
- You have an easy way to roll back changes made if the test fails.
- Both versions of the page/site are showing correctly by using incognito/private browsers.
- The test is rendering properly across different browsers and devices.
- The data is being gathered in the testing software.
You should continue to monitor your test to determine if running the test is having a dramatic negative impact. However, don’t get too afraid if you see a bit of a dip in your KPIs for the test variant/s to begin with. With so little data it’s easy to jump to conclusions and stop a potentially winning test out of nervousness.
If you don’t have the traffic for testing, then using benchmark data (data gathered before you implemented the changes) and comparing this to the data after you implemented the changes can be an option. This way of testing is less scientific as there will be more factors at play, but sometimes A/B tests simply wouldn’t conclude for years due to low traffic volumes.
(Potential) step 5 in your CRO testing strategy – implementation
Let’s say your hypothesis is proven correct (hooray!), now is the time to roll out the change for good. If your hypothesis was not proven correct it isn’t all doom and gloom. Because you went through the steps of analysis, ideation, prioritisation, and hypotheses formulation before you ran your test you will be in a good position to go back to the beginning and analyse why your test didn’t work out the way you wanted it to.
A failing test isn’t a failure as such, it is an opportunity to learn what went wrong. (This can be made easier if you have click and scroll maps of your test). Because you put the effort in earlier on, you can look back and ask; “was our rationale wrong, or was it our solution”.
Not every test will win in conversion rate optimisation, that’s why it’s important to give yourself the best possible chance by doing all the right steps leading up to testing. By following your CRO strategy all the way through, you come up against several points when you can ask yourself “Are we looking at this right?”.
Checking out if you are looking at things right could be; early on if data doesn’t seem to add up across different sources, whilst coming up with different ideas, when prioritising and potentially realising that something you really want to test probably won’t shift the needle, or by writing a hypothesis and realising that the solution doesn’t match up with the rationale.
Starting the process again
Conversion rate optimisation is an ongoing process whereby you learn, test, and learn again before repeating. As long as you are making a record of what went well and why (as well as what didn’t go so well and why) then you will be gaining something valuable each time you move through the steps above.
A great way to build off a winning test is to test something similar across different areas of your site. For example, if showing actual screen shots of a SaaS product worked on one landing page, testing it on another would be a good idea.
Summary
A CRO strategy gives structure to your conversion rate optimisation efforts. If your conversion rate isn’t where you want it to be, or if you are implementing changes on your site without analysis to back them up, then you probably need a CRO strategy.
For a CRO strategy to work, changes must be made one-by-one, or in small, connected groups so you can effectively measure the impact. Making large changes and expecting to see an increase in conversions could backfire if you don’t know what worked and what didn’t.
Everyone in your team should be onboard with your CRO strategy to make sure that tests are being prioritised properly. You don’t want to be testing something just because someone important thinks it’s a good idea. Conversion rate optimisation is about consensus, collaboration, and data-backed decisions over rank.
Want to see an example of a Full CRO Audit? Download it now. Alternatively, why not reach out for a free consultation. We can help your organisation by following the Credo CRO strategy.
Download the CRO Full Audit
Hi, I’m Charlotte. I love writing about user psychology and tips I’ve picked up from leading user research and CRO projects. Get in touch if you’d like to work with me to improve the user experience on your site or click here to learn more about your future CRO team.