Review of Conversion Optimization Minidegree Program (Pt. 7)
What’s the 80/20 applied to Conversion Optimization?
The simple is quite simple, yet most of the times we want to jump to the 20%.
It also happens when marketers and copywriters approach a project (based on experience).
When writing copy, we think that we first need to jump to the writing (the 20%), but that’s far from the truth.
Copywriters need to first do the research (the 80%) to be able to know who the market is, what the market thinks and how it behaves.
The same goes with Conversion Optimization.
Before doing any experimentation like A/B testing, we first need to do the 80%… which is also research.
That’s what I covered this week inside the CXL Institute Minidegree program, along with some ways to optimize a website later on.
Let’s begin.
So from a general overview, you can divide conversion research into 3 parts: site walkthroughs, heuristic analysis and usability analysis.
Warning: If you’re starting a new business, you should forget about optimization. You need is customer development. This simply means figuring out which product to build and for whom… Once you have your first customers and see if there’s a market for it, you can move on to optimizing.
There’s no better way to do this than with the ResearchXL model. It consists of 6 steps:
- Heuristic analysis
- Technical analysis
- Digital analysis
- Qualitative research
- User testing
- Mouse tracking analysis
The best way to start any CRO process is by doing a site walkthrough.
What you’re trying to do is to find any possible issues and places that can be optimized. Sometimes the lowest hanging fruits are cross-browser and cross-device compatibility.
Then, a heuristic analysis is the next step. You want to familiarize yourself with the + and helps you find “problem areas”. Just be aware of your own biases (every single person has them) and sometimes frameworks can help you reduce them.
When you have a clear idea and have examined the website, you can begin a usability evaluation. What you’re looking to get in here is to make websites easy to use so people can browse them without giving it much thought.
Keep in mind that usability is not the same as user testing. Yes, there are some similarities, but user testing has it’s limits. Use checklists to evaluate the site you’re working on.
Moving on, you have the chance to use surveys as part of your research.
Surveys are a great way of knowing the attitudes visitors have towards your website, staying aware of competitors, who your customers are and evaluating new products.
Keep in mind that surveys are a qualitative process, so when gathering all the data you should do it in a zero-sum analysis to then get data from a quantitative analysis.
Although this surveys are done inside the website with popups, you can create surveys that go to customers (not just visitors) and send them via email.
The goal is to get to know those people who just bought something about you, so sending it out the sooner after they bought it is ideal.
As a tip, you want to only ask ope ended questions for this survey so that you can use their language in your copy afterwards.
Another way you an add is by talking to customer support.
You can ask them about the most frequent questions and roadblocks people have when looking to move forward with the product/service. This is going to also help you know if there’s a particular response that tends to work well for a question.
But you can also use live chat transcripts.
You can answer pre-sales questions with it. Reading transcripts from the past 30 days will give you a good insight into what are the questions they have, what pages are they looking for, what product or service are they asking about and what answers tend to work for people to take action.
It’s only after all of these that you can consider User testing. A way to think about it is that when Google Analytics tells you which pages have issues, user testing can help you know “why”. You can learn any places where something isn’t clear, there’s friction or if it’s functional. The true beauty is you don’t need to launch a website to do this as you could ask people based on prototypes.
Lastly, you have mouse tracking. You can find both click maps and scroll maps, which help you see how users navigate your website and whether there’s something that should be added or eliminated.
It’s worth mentioning that you should first conduct a Google Analytics Health check. The last thin you want to do is to being measured something the wrong way or that it’s not tracking as it’s supposed to.
You should be able to understand the current setup that has been created and get areas of focus for this particular analysis.
It’s worth checking things like Funnels, Goal flows (with reverse goal path) and find Key Audience Insights (demographics, behavior), Content reports like landing pages and in-page analytics, and screen resolution, browsers & devices.
What you want is to have clear that everything you’re going to optimize is working properly and always looking for some simple areas that if improving them can lead to a significant increase in conversions.
Lastly, have you wondered if it’s possible to get feedback on the copy you have?
What I mean is, if whether a section of your copy is falling flat or if a particular section is triggering some people to exit the page…
Well. there’s something like that and it’s called Copytesting.
The goal is to give you granular data if your copy is good and where you should improve.
There are programs and services that lets you show your website to a panel and ask them questions about particular sections. This is better in some instances than A/B testing because you can know they why it’s something happening.
And you may be worndering about how to do all of these analysis inside your website.
That’s why Google Analytics is your best friend along this process.
Some of the first things you want to check inside GA is evaluating the traffic quality you have. Plus, focusing on metrics that matter like bounce rates is important, but it’s also based around context. For instance, a 90% bounce rate in a landing page is not that bad when you have a 10% conversion rate. You can then start seeing how to improve it of course, but keep in mind that it’s all contextual.
To get more insights from you market, using tools like secondary dimensions and advanced segments will get you more data on it. The best part is that you can also build advanced segments yourself based on what you are looking for, like comparing mobile vs. desktop.
Laslty, you have event tracking. It’s a way to to track user interaction within your page (a good way to deal with the misleading numbers of bounce rates as well).
That was what I learned in the 2 courses from this week.
I find them very detailed enough with a clear process to go after when optimizing websites or landing pages. These courses were very insightful on how to have a framework that let’s you as an optimizer understand what’s working on a website and where you can improve it.
Next week, I’ll be talking a little bit more about Google Analytics in great detail so you can know more how to use it in your conversion optimization process.
Ivan I.
