My Mother is having her home re-modelled. Her builder was recommended by a friend (word-of-mouth is the best marketing method around) and he has no online presence, as such, she offered up my services to build him a website.
Which meant that I had to break it to her that I am no longer a website designer & developer. Frankly, the idea of building a website from scratch fills me with dread, despite the many frameworks and preprocessors that have emerged since I left pure front-end development. It’s an uphill struggle and it’s why I much prefer the fast paced test & learn life of CRO.
But, it meant that I had to explain to her what I do now, and it turns out, I do quite a lot.
I’ll be the first to admit that my analytics skill-set is my weakest in regards to CRO. I have no data science background & the nomenclature of the analytics world confuses me, but I can still find my way around a Google Analytics dashboard.
I lean on analytics teams to supply me with the data I need, but it is still my responsibility to know what to look for. Typically I start with broad funnel reports, but eventually you need to tunnel down into landing page conversions, audience segmentations, bounce rates and understand trading peaks & user journeys.
Having a good relationship with the analytics department is vital. Without their insights I would be clueless about where to begin improving a website. But being able to generate reports, or even just make sense of analytics dashboards yourself enables you to monitor your test successes (and failures) without impeding colleagues.
I have a long journey ahead of me to improve my analytic knowledge, but understanding the basics is a good start. Maintaining good working relationships for those times you need assistance currently works well for me too.
Working in CRO requires constant test idea generation. You should definitely be using your analytics data as a starting point but establishing what exactly can be improved on a webpage is not cut & dry. Luckily, a multitude of tools and methods exist to help pinpoint what aspect of your page is failing to convert.
Some of the techniques I used for research include:
- Heat maps
- Visitor recordings
- NPS & on-site surveys
- User testing
- Customer Support feedback
Knowing when and how to use all of these techniques is a skill in itself, but being able to correlate all the relevant data into one place and then come up with a valid test hypothesis has taken years of practice.
But your resident CRO expert should not be relied upon to generate all the ideas — your colleagues will undoubtably have loads too, even if many of them are subjective. Running workshops and turning vague ideas into fleshed out tests can make sure the whole business feels involved in optimisation.
Prioritisation and Road-mapping
Deciding what tests to run & when required its own article, and yet, putting together a road-map of tests is often overlooked as a required skill. The number of considerations you must juggle when creating even a short term road-map is mind boggling.
- How high a priority is the test?
- Will the test clash with website releases?
- Will the test clash with other tests?
- Is the test due to run alongside marketing campaigns?
- As well as considering any trading peaks, content dependancies, engineering bottlenecks, or HiPPOs.
It’s a minefield! And it’s why I create new road-maps every 3 months because business priorities might change, website releases might alter the codebase and test results can alter your perceptions.
Test Building, QA & Monitoring
My background as a front-end developer was how I initially got into A/B testing. This means that I can actually develop the tests that need to be run.
QA-ing tests can and should be done by anyone familiar with the website. Finding bugs before a test goes live is vital and although not all CRO specialists are required to build tests themselves — that is what teams and agencies are for — understanding the code enough to fix a bug is super helpful!
Personally, I love digging into existing website code to see how it was built and how it can be manipulated. It’s another way for me to generate test ideas.
I also spend a lot of time keeping an eye on tests which are running. Often you’ll have multiple tests at various states of completion, all of which require monitoring for any worrying behaviours (in which case it might be time to stop the test) for bugs (in which case, stop and fix the bug) or monitoring for completion (in which case, consider whether to stop the test or to push to 100%.)
Once a test has run (or realistically, whilst a test is running) I’ll be checking uplift (or downlift) and comparing it with analytics data to ensure there are no discrepancies. Putting together an easy-to-understand report which can be shared throughout the business is another part of a CRO’s job. There is a lot of documenting involved.
Often, you’ll also be required to present test results to rooms full of managers and stakeholders. Public speaking is a skill I’ve only just started working on, but you must be able to put together an engaging and informative presentation as well as speak clearly & know your material.
Communication is a key aspect of maintaining good working relationships as well as encouraging a test & learn culture.
So there you have it, the main responsibilities of being a Conversion Rate Optimisation specialist (or an Optimisation Team.) I’ve not even delved into some of the challenges working as a contractor bestows upon my life. Nor the sheer amount of learning and research that goes into keeping up-to-date with industry standards, learning new tools, and generally staying knowledgeable.
It’s a demanding job with lots of balls to juggle, but then again, I also know how to juggle, and I love it!
What skills does your role require? Has reading this made you realise that you do more than people expect? Let me know!