CRO (Conversion Rate Optimisation) is not just for large traffic sites with whole testing teams in place…
In this post, I want to talk through some challenges of CRO for SMEs and low traffic sites, why CRO is so important for business growth (no matter the size of your business), and how the CRO landscape is rapidly changing in the marketing world.
The challenges of CRO for SMEs:
- Having enough traffic to test on.
- Investing in conversion research to identify test opportunities.
- Knowing what to test first.
- Having internal resources and deciding the right technology stack.
Conversion rate optimisation is more than changing a word in your top fold or the colour of a button.
Often testing these small changes on a site wastes valuable testing time and user interaction with your site. Tests should hypothesise real change in user experience and behaviour that helps drive your North Star Metric (mainly top-level business goals such as revenue or profit).
Do you have enough traffic to run test programs?
The bigger the sample size in tests the smaller the margin of error. This means the more site traffic you have the better as you can run tests in a shorter period of time with high confidence in the statistical significance of the final result. The faster you run tests the more tests you can run per year and therefore the more agile you can be in optimising your site.
But what about sites that don’t have a lot of traffic per month?
This shouldn’t stop sites from constantly experimenting with site navigation, user experience, and overall messaging. It just needs to be done slightly differently from the higher traffic sites.
Conversion research is an absolute must! This will allow you to find the right areas to focus on to make sure your time is always well spent (more on this in the next section). Sometimes, with smaller traffic sites, the only way to test a change is to change it, put it live, and closely monitor user behaviour on your site. As a rule of thumb, any increase in conversions less than 15% will be hard to attribute to any one particular change you have made. But with slow and steady changes comes slow and steady site conversion increases and you can start and become more and more confident in your site changes being the cause for the business needle moving forward.
That’s not to say you can’t run tests, but be prepared to run less of them for much longer to see statistical significance.
Investing in conversion research
Conversion research helps you identify where your site is letting you down. But why does it have to be driven by data? Why can’t you just start testing things you think aren’t working?
Naturally, everyone has their own cognitive biases. Everyone thinks differently and has opinions that don’t match the next person. Data and science remove cognitive bias in all research. You and any stakeholders you may report to, can’t ignore or debate the validity of facts and objective evidence!
Sources of insight:
- Google it! – Oh yes, that’s right. If you’re not able to invest budget into conversion research then Google your problem and see if anyone has come across the same thing. Chances are, someone has already tested it and there might be a case study for you to read. HOWEVER, not all tests are the same and not all audiences behave in the same way so you need to take all historical tests you’re applying to your own site with a pinch of salt!
- Existing research – Do you have any customer personas or historical tests that have been run on your site before? What can you tell about user behaviour based on old tests?
- Analytics – Analytics lets you see if what you think is an issue is an actual issue with visitors before your start delving deeper into customer research.
- User testing – These can be moderated (focus groups, usability testing etc,) or unmoderated (heat mapping, eye tracking, and screen recordings). These help you understand what users do.
- Qualitative research – This helps you understand what users/customers say/think. Mainly made up of 1:1 customer interviews, questionnaires, website polls, and reviewing live chat and call transcripts. These give you much more depth and understanding into your customers values, fears, uncertainties, and doubts (FUDs) to help you with the correct value propositions and messaging.
Overall, you shouldn’t be making decisions on what you believe, but what your customers/clients really want and say by way of collecting as much useful data as you can before you make big business decisions.
Prioritising and managing test ideas
After you’ve conducted all your conversion research you’ll have a list of problems your data is showing that can be a potential conversion blocker on your site. But how do you start and prioritise these problems, and what changes are needed to solve each problem?
Usually, without any prior organisation or plan of action, many people test what is at the top of their list and work their way down. But what’s at the top of your list is only the first problem you found, it doesn’t mean it’s the most important or going to have the biggest impact on your North Star Metric!
There is a much more efficient way to categorise tests:
- Just do it – These are no brainers. These are serious problems with your site that just need fixing asap, such as a broken checkout page or a landing page not loading on a key browser.
- Need more information – Sometimes you don’t get all the information you need in the first round of research. These are problems you’ve identified but you still need a bit more information to guide you to understand how to fix it. This could be conducting more 1:1 interviews with customers but only addressing a particular part of the site.
- Hypothesis – This is where you’ve found a page, widget, or process that’s just not working well but we don’t see a clear single solution. You need to brainstorm hypotheses that could have an impact on this part of your site.
Hypotheses can be further prioritised using a range of models (ICE, PXL etc) that is ordered in the possibility to have the highest impact for the lowest effort in time and resources.
If you don’t prioritise tests appropriately then you risk spending time on tests that are only likely to move your KPI by a few %, and not the “low hanging fruit” that can have big impacts quickly, or the tests that might take a little bit longer in time but delver a much bigger increase in your business targets.
Internal resources and external technology
What tools do you need to run CRO and research?
Analytics – Google Analytics, Adobe, Mix Panel, KissMetrics…
Survey Tools – Google Surveys, Survey Monkey, Qualtrics, Qualaroo, Hotjar (on-site surveys and polls)…
User Intelligence – Hotjar, Crazy Egg, Mouse Flow…
A/B Testing Tools – Google Optimise (free), Optimizely, Omniconvert, VWO,
Creating a technology stack that allows you to collect and analyse a range of data can start to become costly when you begin to add tools to your inventory. But agencies, such as ourselves, already have subscriptions to many of these tools to use across a wide range of client needs.
Historically, businesses would also need a few members of staff to plan and run a testing program normally made up of a CRO specialist/researcher, designer, and developer at the minimum.
The rise in no-code landing page builders, such as Unbounce or page builders that can be installed on WordPress means that marketers with little coding skills can build no-frills pages that are relatively responsive across all devices.
Whereas this helps smaller businesses because they don’t need additional resources and developer time can be spent on other projects, it puts a lot of strain on the present-day small business CRO (normally a one or two-person team in much smaller companies) who now works across all skills and pillars of testing from research to wireframing, to designing and building the landing page(s) variants, to analysing results.
The only problem is that although they can work across all pillars of CRO, they only have so many hours per day to work. So increasing your CRO tests means investing in growing your team and technology, or outsourcing a part of your testing program to a dedicated agency.
In house testing team or outsourced?
One major advantage of outsourcing conversion research is that there is a lot less bias in the research. Agencies have no former opinions or knowledge of your business/customers and are therefore creating a research project from scratch without bias. This gives you a research report based purely on data, evidence, and customer interview analysis, instead of historical business decisions and product reviews.
Having a completely new person audit your site and talk to your customers gives it a fresh pair of eyes and often some quick wins on-site usability as the researchers become the first tester in initial site walkthroughs.
What happens if you make changes without conversion research and proper testing?
Without conducting data and conversion research you end up making decisions based on opinions (usually the highest-paid person, unfortunately) and gut instinct.
You don’t want your tests to be political, force of will, or a good idea to everyone. They need to be backed by data or you risk wasting time and resources running in the wrong direction.
What’s worse, is basing all your decisions on what your competitors do! A bit of competitor research is vital, but too often we find ourselves thinking “oh well this must be working for them as they haven’t changed it”, or “this is the best way of displaying information because I like the way it looks”. When in reality, we have no data or access to their analytics to back up those claims. Instead of looking at a competitor’s site to steal ideas, we should be thinking about the differences/value we bring customers and how we can portray that in a language they speak and understand.
By stealing ideas we start to play it safe (stock photos, corporate layouts, and bland messaging). And safe isn’t revolutionary.
Data-driven evidence shows you exactly what is stopping people from converting on your site and gives you an objective list of changes to test without succumbing to the “helicoptering effect”.
This is the “panic mode” of CRO programs (when tests are failing and goals aren’t being hit), and EVERYONE with an opinion gets involved across the business. It starts to confuse programs, know what’s running and what’s not and also leads to changing things before tests have been running long enough to get a statistical result.
Interested in conversion research?
Combing research and strategy together we have launched our new Blueprint service…
Market research and strategic marketing have been combined to give you an actionable research Blueprint. It delivers valuable customer, competitor, and conversion insights and translates these into a detailed marketing strategy or test optimisation program to start maximising profit. All based on extensive field and desk data-driven research.