When discussing Conversion Rate Optimisation I like to challenge people to answer the question:
“Can a small decrease in conversion rate still lead to a positive test outcome? And if so, in what circumstances?”
It doesn’t seem intuitive but once people think about it they usually realise that there are scenarios when other metrics are equally or more important. How many scenarios can you think of? (if you get more than three you’ve done well, my answers are below).
If this is the case, how important is Conversion Rate as a metric? Are we too transfixed with Conversion Rate?
Why conversion rate is a great metric
The simple equation of how many people who walked through the doors of our virtual stores were we able to sell to/sign-up/etc. is a great metric to review. There’s no doubt that increasing the number of people we can convert into customers is important. From an analysis point of view it is good to measure and analyse ratios rather than just raw numbers. If the scale changes, the ratios should still be a relevant indication of your ability to serve customers.
Why conversion rate is not enough
Understanding conversion rate is important, it’s a valuable metric but it needs to be considered in context. Conversion rate is widely available in many of the tools that we use and this makes it more attractive than other metrics which can be harder to record and measure. Businesses that we start working with often do not have a clear picture of returns or lifetime value within their primary analytics tool.
It’s important that we work to close holes in our data used for analysis, rather than simply accepting the metrics served to us.
My answers below indicate some of the scenarios in which conversion rate alone would give us an incomplete view of performance.
- When Average Order Value increases, you may have a more profitable variation.
- When the product mix changes significantly to favour higher margin products.
- When there is a considerable saving for the business. For example if you are producing content that requires a lot of effort, how much is that effort worth in terms of conversions?
- When return rates are reduced. Track return rates whenever possible.
- When there are positive changes in Lifetime Value or Purchase Frequency you may improve profitability.
- When there is a significant learning that leads to further testing.
I’ve been building up this list, but any ideas in the comments would be very welcome.
In summary, conversion rate is a really valuable metric, but don’t use it in isolation. We need to review profit, margins, returns, internal resources, lifetime value and more.