When calculating impact at Persado, we always want to use a metric that is not in the optimization funnel as the denominator, like impressions. This article walks through why.

A marketing impression is one opportunity for someone to see your message. That’s it! It does not mean they:

It only means the marketing content was shown to them—in other words, they’re the audience that receives your message. Impressions are considered top of the funnel because they constitute the very first step.

Real world example: If 100 people walked past a store window on a given day and saw the sales advertisement, that would equal 100 impressions.

We want to use impressions as the denominator for calculations of click-through rates (CTR), conversation rates, and revenue per email. This is because these metrics are about behavior, and we can only judge behavior fairly if we count all opportunities, not just consumers who already showed interest, which would introduce bias.

Here’s an example using the below correct formula for the CTR for an email, assuming impressions as the denominator. Let’s assume we had 10,000 impressions and 500 clicks. This results in a 5% CTR, meaning that for every 100 people who saw the email, 5 clicked.

If we were to calculate CTR incorrectly, we might use a denominator like opens—which would result in inflated, misleading data as shown below.

Why is this incorrect? You’re ignoring the 8,000 people who saw the email but didn’t open, and you’re only measuring the already-interested audience (i.e., you’re cherrypicking!). This inflates performance and hides problems like:

Let’s get even more technical.

Using a denominator that’s something that Persado optimizes for (i.e., opens/clicks) creates a misleading response rate and lift that can be very easily misinterpreted. The reason? You’re comparing two fractions where both the denominator and the numerator are being affected simultaneously. As a result, the final number would only be a result driven by which KPI (numerator or denominator) was affected the most.

Consider the below example. Let’s assume both the Persado message and the Control message receive 100 impressions.

As you can see from the table, Persado is performing better both in opens and in clicks since it has a greater response rate in both cases (impressions is the constant denominator).

But…if you were to go through and calculate the Clicks/Opens metric, it seems as if suddenly the Control message is performing better and we have a -20% lift—which is not true.

This happens because the 50% uplift in opens (numerator) is larger than the 20% uplift in clicks (denominator).

In summary, the Clicks/Opens metric should not be used since the only thing that impacts it is which KPI lift was greater than the two. The correct way to independently use the Open and the Click Rate, as discussed in the basics above, is to express what Persado's lift is in terms of opens and clicks.

Impressions represent the largest, fairest opportunity pool, making them the ideal denominator for most marketing metrics calculations measuring overall effectiveness. Anything smaller (like opens or clicks):

Keep in mind, though, that impressions aren’t a “one size fits all” denominator. Read our comprehensive guide to marketing metrics for guidance on calculating other metrics.