Everyone looks at global metrics.

Otherwise we wouldn't have the "our conversion rate is X and we want it to be Y conversation" a few times a week.

It is a baked in part of the equation in a complex space like ecommerce.

The challenge we put to the brands we work with is, how you can use these metrics in a way that lends confidence that you are focusing on the right thing?

For example, we talk quite a bit about PDP → Add to Cart rate.

Having a site strive for 12% is vital for continued ecommerce growth, based on standard cart abandonment levels.

But a 12% PDP > Cart rate can look quite different when you analyze at the channel level.

And the path to fixing a lagging site, can change entirely based on those underlying numbers.

We expect high-intent sources like direct and email users to come in well over that 12% rate for most stores. Organic can be a mixed bag. And paid, by design, should be a fair bit lower.

Here is what a 12% PDP → Add to Cart Rate can look like by channel in a hypothetical scenario.

The distribution above would be a standard example of “good” performance.

High-intent channels are doing what is expected of them. Paid is lagging a bit, but still in an acceptable range.

Good performance is easy enough to assess.

Let’s look at two scenarios where global site PDP → Cart rate is underperforming at the same level, but channel figures are very different.

Store 1 - High-Intent Channels Underperform

Store 2 - Paid Channels Underperform

If you just looked at the top level PDP → Cart rate you would see the same rate and probably not think much of it.

However, the realities for these stores are so far apart.

In Store 1, channels that should be warmer are simply too low for our liking. This suggests that there may be UX issues, or a general miss on product market fit.

→ We would try to smoke out things like responsiveness quirks, out-of-stock issues, or browser-specific issues first.

There might be more at play in the market itself, but, we like to focus on controllable factors first.

In Store 2, the high intent channels do a bit better, but paid traffic dips into a danger zone. This store is less likely to have UX friction, and more likely to have a big disconnect between the pre- and post-click experience (causing poor paid performance).

→ We would zero in on the paid channels and try to close the gap between the pre-click ad expectations and what the user sees when landing on site.

Post-purchase surveys and customer reviews mining are likely needed in order to help understand why things aren't clicking.

Either way, the path to fixing performance on these stores is very different - but you would never know that by only looking at global metrics.

Every data point tell us something. The more you can learn to infer the action required to improve it, the more confident you will be that you are doing the right thing as a result.

‍