Skip to main content
Coherence logo Coherence logo

Articles /

Analytics should not be confused with insight


Mar 2018

Molly Peers

Head of Insight

The stark contrast between analytics and insight

It’s easy to say, but there is a stark contrast between analytics and insight - and this seems to be the little sprinkle of magic many companies are missing.


"All I have so far is a report."

"I have 300,000 visitors. Excellent! 900 transactions. Great! I changed a button and now it’s getting 25% more clicks every day. This is all good news. Let’s make a dashboard."


"Well I can see that things are good. I can walk into my boss’ office and say 'everything is brilliant'. Now, if this is my only objective for my website, then this is great. If, however, I’m embarking on an endless optimisation journey, I may want to ask one simple question."


This is so frequently overlooked, but it shouldn’t be. In fact, it can be severely damaging. Let’s play this out.

You have two websites: one is a video subscription website, and one is a website that sells language courses. They both have exactly the same call to action.

On your learning website you change the words on your primary call to action from 'Sign up' to 'Learn more'. Your conversions go up. Yay.

You do the same to your video subscription company. Conversions go down.

Moral of the story:

Understanding 'that' it worked is one thing. Understanding 'why' it worked is entirely different - and that is where the gold dust is.



Now, I appreciate that most people will be shaking their heads saying 'well, duh!', but take a moment.

How many numbers have you churned out in the last week for various people? And how many of those have been appended with a 'why this happened?' I’m guessing very few.

Analytics tools, supercomputers and even vacuum cleaners are getting pretty smart. However, they are still no match for the human mind. The human brain is said to perform 10,000 trillion calculations every second. This is why we are far better equipped to cope with the complexity of 'why'.

Let’s use it, before our vacuum cleaners become self-aware.

3 Tips for Reporting "insight":

Beware of confirmation bias.

You thought the green button would work better because it’s more friendly. The green button does better. Must be because it’s more friendly, right? Maybe, but remember to ask yourself, is that really why it did better? Or is it because your website is orange and it has the highest contrast against it?

Don’t be afraid of theories.

If you are lacking in a psychology, economics, meteorology and web design degree, you’re not alone. Present your 'why' as a theory. It doesn’t have to be concrete, but it shows an understanding and consideration as to the bigger picture. Remember: context is king.

Back it up with data.

The best part about insight is that you can usually find other data that supports your insight. Take, for example, the green button theory above. It may be that you have a webpage that is blue and the button has not done so well there, further enforcing your theory that contrast was the driver.

Want to understand your data a little more? Why not get in touch!

Molly Peers

By Molly Peers
Head of Insight

Share this