Is Facebook wrong to conclude that 5 ad impressions drives optimal impact?

George Paton-Williams, Head of Marketing

Whenever I see a new piece of research that recommends the marketing industry to do one thing or another, I’m normally pretty sceptical. Usually it rather conveniently fits the agenda of the company doing the research, or the sample is either too small, too niche or just simply irrelevant.

But when I read last week that Facebook had done a piece of research about performance at different ad frequencies I thought I’d read on.

Sure, it is pretty standard topic that has been tackled before. But surely Facebook would run a pretty solid piece of research with a hefty sample size and some valuable results and insights. It would be interesting to get Facebook’s take, I thought.

Facebook were looking to calculate the increase in what they called ‘desired response rate’ (a combined percentage increase in action intent and ad recall rates). They would look for the point at which the increase in ‘DRR’ plateaus to determine the optimum number of impressions before returns diminish.

The answer, Facebook lands at, is 5 impressions.

Reproduced from, and copyright of, Facebook IQ
(Although why, based on that graph, they land at 5 impressions rather than 6, I’m not too sure.)

However, you could argue that this graph shows that the optimal number of impressions is not 5. In fact, it shows you that Facebook’s research suggests that capping impressions to just 1 is the most effective way to spend your budget.

Here’s why:

The chart above shows that the average ad on Facebook can increase action intent and recall to a maximum of 26% after 8 impressions.

The first impression contributes about 18% of that 26%. That is over two-thirds of the advert’s total impact. In fact, 69% of a Facebook ad’s ability to drive a response comes from the first time a person sees it.

Ad impressions DRR (%) Incremental contribution % total 26pc DRR Cumulative % of total 26pc DRR
0 0 0 0 0
1 18 18 69% 69%
2 21 3 12% 81%
3 22 1 4% 85%
4 23 1 4% 88%
5 24 1 4% 92%
6 25 1 4% 96%
7 25 0 0% 96%
8 26 1 4% 100%
There is something uncanny about the stat that 69% of an ad’s work is done on the first impression. Marketing academic, John P. Jones, once showed back in 1990 that “the first opportunity to see [an ad] contributes 73% of the short-term sales effect of advertising”.

Pretty close, huh?

Then, the 2nd impression contributes the next 12% of the remaining impact. So when you combine the impact of the 1st two impressions, 81% of the ad’s work is done.

While showing the same advert a further 3 or 4 times might take you to the entirety of 26%, following this advice will actually hold you back from your overall aim of earning as many new customers as possible.

Let’s play out a few imaginary scenarios where you only have enough money to spend on Facebook to earn 100,000 impressions.

Scenario one – frequency of 8
You hit everyone with the advert 8 times. You have therefore reached 12,500 people. And at a ‘desired response rate’ of 26%, you have earned 3,250 desired responses.

Scenario two – frequency of 5 (Facebook’s optimum, remember)
You hit everyone with the advert 5 times, reaching 20,000 people. At the DRR of 24%, you have earned 4,800 responses.

Scenario three – frequency of 2
Two exposures each means a total reach of 50,000. At the DRR of 21%, you earn 10,500 responses.

Scenario four – frequency of 1
Your reach of 100,000 people at a DRR of 18% means you walk away with 18,000 responses.

So while the likelihood of impacting any one person drops as you reduce frequency, you are able to spread your budget across more people (i.e. increase reach) and drastically increase the absolute number of people who can respond to your advert.

And ultimately, a larger absolute number of outcomes is more important than small incremental increases in percentages.

I don’t have the data to support the following suggestion, but I imagine that response rates are not a constant across companies. For example, if your industry has a long purchase cycle (perhaps like insurance or automotive) you would maybe benefit from greater frequency.

Also, reach might not be too much of an issue if your budgets are vast enough.

But for most companies, Facebook has just told us that one or two exposures to an ad is enough. Invest the money you would normally spend on those 3rd, 4th, 5th impressions (and so on) elsewhere. Invest it in maximising your reach and producing more and better creatives.

Scale up your performance marketing

7 steps to growing your app UA profitably

Get UA insights straight to your inbox


Linen Hall
162-168 Regent Street


+44 203 096 5108

How can we help?