
What are the steps to conversion optimization wisdom?
How do you go from looking at a website which works poorly to knowing what to do to make it work very well?
Get the data
You need to start with data. Without data you have virtually nothing at all.
Steps:
- Install Google Analytics
- Configure Google Analytics properly (everyone makes mistakes here, or even worse – ignores it completely)
- Install Hotjar for heatmaps, click maps, scroll maps, session recordings
Turn data into information
Now you start gathering data. Gradually, it turns into information. You start making sense of it.
Now it’s not just the number of conversions divided by the number of visitors. You start seeing segments, different devices, seasonality.
You now have all the building blocks to start extracting knowledge from this vast amount of information.
Helpful
How to Build A Google Data Studio Dashboard – Step-by-Step Tutorial
Turn information into knowledge
Information is useless if it doesn’t give you knowledge.
You need to be able to make sense of all the data.
For this, you need to filter what’s useless and to focus on what matters.
Steps
- Know which Google Analytics reports are essential and provide valuable info
- Know which reports are just background noise and don’t provide any insights
- Use dashboards to start seeing the trends
- Set up Google Analytics alerts to learn about important things and changes as soon as they happen
- End result: distilled, quality info extracted from the sea of chaotic data.
Helpful
8 Must-Have Google Analytics Reports for Ecommerce Optimization
Great Google Analytics Alerts: Less Time Panicking, More Time Solving Issues
3 essential analytics insights for all types of businesses
Turn knowledge into insights
Now you start seeing how the elements interact.
Which parts are independent, which ones correlated, which parts influence which other parts, and in what way.
For the most part, you now have the answer to “what?”. What doesn’t work? Where? In what way?
Now it’s time to figure out why.
For that, you briefly step away from qualitative (data) and step into quantitative (people).
Steps
- Run user surveys (email, website)
- Run user testing (with 5 people is enough)
- Watch user session recordings (mostly boring but sometimes eye-opening)
- Start seeing the trends
- Keeping in mind everything you saw in analytics, see the connections between what data tells you and what users show you
Now you know not just the “what” but also the “why”.
Helpful
Why is my conversion rate dropping? 8 common reasons
Turn insights into wisdom
When you know what doesn’t work (data told you) and you saw clearly why (people showed you), it’s time to go from “doesn’t work” to “works very well”.
Steps
- Take what doesn’t work
- You know why it doesn’t work, from the previous step
- Fix it in a way that seems most effective to you
- Test it
- See if it’s the most effective to all the visitors, not just to you
- If it worked, fix the next thing
- If it didn’t, find another way to fix it, and test it again
- Gain experience
- Run better tests
With time and experience, the yellow road from the picture above will be more straightforward, and the time to get to the best results will be shorter.
Conclusion
Steps to go from “doesn’t work” to “works great”:
- Set up Google Analytics and all the other tools. Get lots of data.
- Start getting information. Start making sense of it.
- Gain knowledge based on what the information provides you. You now have all the reports, alerts and dashboards to see the big picture.
- Start seeing trends and dependencies. See what works, what doesn’t. Learn why, directly from people.
- You know how to get from “doesn’t work” to “works great”. You test it. You learn from it and keep repeating the process.
It’s important to keep in mind that if you miss any of the steps, you’ll be severely handicapped. It’s hard to get great results based on “garbage-in-garbage-out”, even if all the subsequent steps are performed perfectly.
PS. To avoid “garbage-in-garbage-out”, be especially careful at the data-gathering and analyzing steps.