How to Research and Prioritize Growth Experiments
You may very well be sold on the idea of running experiments to improve different conversion metrics but now the question inevitably is, where to even start?
To help figure that out, you should always keep in mind that conversion optimization is a process. It’s a systematic way of finding opportunities for growth and developing data-based ideas for how to build upon those opportunities.
This process will help show you where/what problems need to be addressed, how to turn those problems into hypothesis, and finally how to prioritize testing of those problems to try and fix them.
Research
To start discovering what problems need addressing and where to find them you need to look at the data.
Four different types of data for you to start analyzing are:
- Technical
- Heuristic
- Digital
- Mouse tracking
Technical Analysis
During this analysis, you will be looking to see if there are any underlying issues with your site’s performance. This could be in terms of broken pages, bugs, site speed, etc.
A quick audit you can do to spot potential issues is to look at your Google Analytics and ask yourself “does anything seem broken or underperforming?”. You can view how different browsers and devices are converting and see if there are any out of the norm.
If you see that 10% of consumers using Safari or Chrome purchased your product/service but only 1% of consumers using Internet Explorer purchased, that tells you there could be some issue with your site on Internet Explorer and you need to go deeper to see what that is. The next step would be to either go on your site using Internet Explorer and walk through the customer buying journey to see where the hiccups are or view a consumer’s interaction with your site using specific tools which I’ll go over in a bit.
Doing a technical analysis can show your where the low hanging fruit is in terms of fixes that will provide instant results.
Heuristic Analysis
This is used to understand a user’s experience on the website. To perform a heuristic analysis you will go through the entire site and critique every page based on fixed criteria. This criterion should include aspects like clarity, relevancy, friction, distraction, and motivation.
- Clarity - Is it perfectly clear and understandable what’s being offered and how it works? This isn’t just saying is your value proposition clear. You need to make sure consumers understand your pricing, features, product pages, etc.
- Relevancy - Does your website related to what the visitor thought they were going to see? How frustrating it is when we click on a link expecting a certain product or answer to our question and get nothing close to it.
- Friction - Find all the sources of friction on key pages. This could be like difficult or long processes, not enough information, bad UX, fears about security, etc. We’ve all been on a website before that look sketchy and think to ourselves “no way I’m giving them my credit card information.” Don’t be that site.
- Distraction - Look out for anything that deters visitors from taking the desired action. This could look like animations or sliders stealing attention, providing too much unrelated information, irrelevant page links bringing visitors farther away from purchasing, etc.
- Motivation - What is being done on-page to increase motivation for visitors to take action? Is there some sort of urgency factor? Is the sales copy persuasive? Is there enough product information?
As you go through the website, write down everything that you notice regarding the above areas so you can then go back to it and figure out how to test and make better.
Digital Analytics
Your digital analytics is going to provide so much information on your customers that you might not even know where to begin. I’ll be going over analytics in great detail in another article but for now, here are some major jumping-off points to get you thinking about using digital analytics.
A good place to start is to find the “leaks” or where people are dropping off in their customer journey. Are they coming to the home page and leaving? Making it all the way to check out and then stopping? These are all red flags that would require further investigation into whats going on.
You can also segment your visitors by areas like devices, demographics, geographics, etc to see who is converting higher and lower than average.
By tracking what users are doing on your site you can see what actions correlate with higher conversion. For example, how great would it be to know that somebody who goes to a certain page on your site converts 12% higher than someone who doesn’t? You could then try directing your traffic to that page or see what’s on that page that’s improving conversions and replicate it across the site.
Mouse Tracking
You can use tools like heat maps, click maps, scroll maps to actually see what actions users are taking on your site. This data helps show you what’s catching visitors’ eyes and taking their attention, where they are clicking, how far they are scrolling on your page, as well as what they want to be doing.
For example, if you have a picture on your cooking website of a recipe and you see that users are clicking on it and going nowhere, you should try making it clickable and directing them to another page that has a call to action on it.
Prioritizing Tests
After all this research you should have an extensive list of problems to improve on. But how do you figure out which to start with?
The first step is to categories all of the problems into groups like:
- Instrumentation - something isn’t getting measured correctly
- Just do it - some issues you can just fix right away like a broken link or something like that
- Test/Hypothesize - you need to test and see what the result will be of the change
- Investigate - you need to further investigate the problem to see what is the underlying cause or how to fix it
Once your problems are grouped, you can then start to rate them. A simple way to do this is to use a 1-5 star scale.
- ★★★★★ - This rating is for a critical usability, conversion or persuasion issue that will be encountered by many visitors to the site or has high impact. Implementing fixes or testing is likely to drive significant change in conversion and revenue.
- ★★★★ - This rating is for a critical issue that will be encountered by many visitors to the site or has a high impact.
- ★★★ - This rating is for a major usability or conversion issue that may not be viewed by all visitors or has a lesser impact.
- ★★ - This rating is for a lesser usability or conversion issue that may not be viewed by all visitors or has a lesser impact.
- ★ - This rating is for a minor usability or conversion issue and although is low for potential revenue or conversion value, it is still worth fixing at lower priority.
The two most important criteria for when giving a score is going to 1) the ease of implementation (time/complexity/risk) and 2) the opportunity score (basically your subjective opinion on how big of a lift you might get).
There are also more accurate but complex ways of prioritizing tests like CXL’s framework which they provide for you free on their site.
If you’re interested in learning even more in-depth ways to research and prioritize tests I highly recommend checking out their website.