What is an effective frequency for advertising?

By Nigel Hollis

Let’s be clear up front, the answer to the question posed in the title is it depends. The frequently used rules of thumb will be frequently wrong because the specifics of advertising objective, media channel, and product category make them so. But there are guidelines to be observed and there are myths to be ignored.

Advertising needs to influence people

Over the years we have learned a lot about the way that people respond to advertising, both for traditional media and for digital. Unfortunately, many practitioners, be they advertisers, agencies, or part of the media ecosystem, seem to have forgotten some, if not all, of that learning. The fundamental problem appears to be that digital practitioners often forget that the ultimate objective of advertising is to influence people’s behavior at scale. And, no, Android is not a person.

Prioritize effectiveness over efficiency

The first thing to remember is that effective advertising is not the same as efficient advertising. And effectiveness should always come before efficiency. Reaching as many people as your budget will allow is more likely to drive sales than limiting reach to a subset of high potential buyers. But if you can reach a high-quality lead, it may well be worth paying over the odds to do so.

Mindlessly seeking the lowest cost per click is a recipe for ineffective advertising, because, as a smart person once told me, the first dollar you spend will always be the most efficient, but it will not achieve much on its own. And yet, many people seem intent on optimizing “cost per” metrics, while completely ignoring the scale of response to advertising.

Frequency rules of thumb

When I started work the primary media channel was TV and the basic rule of thumb for frequency was three exposures. No one seemed to question over what time frame those exposures were meant to be delivered. And was it three because that was the average number of exposures before someone got to watch the ad or was it related to sales?

Now, the rule of thumb for digital appears to be around seven. Both may well be roughly right for an individual in a short period of time. But both will vary substantially depending on the specific case. What I do not believe is true is the prevalent belief, particularly among B2B advertisers, that digital advertising requires far higher frequencies than seven. And I find the idea that you would prioritize frequency over reach just weird. To explain why, let’s start with the basics.

Diminishing returns

All advertising suffers from diminishing returns to frequency. The incremental response to the first exposure is always the highest and the incremental response diminishes from there on until it plateaus. Most of the cumulative effect comes from the first few exposures. This post from Facebook IQ shows some typical response curves, although I do note that it is based on attitudinal not behavioral data. I have seen similar curves based on behavioral data, and they do not look too different, it is just that the response is a lot smaller. Whether it is worth continuing to expose someone really depends on why they have not responded yet.

Good creative matters

The Facebook IQ post suggests that, “‘good’ creatives—those with higher performance—tend to garner better results for more impressions compared with other creatives.”

I’m going to be pedantic and correct that statement because good creative gets a better response at all levels of frequency. From the very first impression it is obvious which creative is good and which is not. Good creative is likely to plateau sooner than bad because it reaches and influences people likely to respond quicker, contrary to what is written in the post. The poor creative has not plateaued. It will likely continue to deliver a tiny fraction of what the good creative achieved per exposure for many more impressions. And that is the key takeaway for people who believe excessive frequency is the key to advertising success, if your creative sucks no amount of media tail is going to wag a dead dog.

Three reasons for a poor response

Assuming you got your targeting right, there are probably three reasons why digital creative might fail to generate a good immediate response.

1) Your ad did not earn attention

Let’s say you have a carousel ad running on Facebook, Instagram or LinkedIn. People will likely notice your ad, but whether they will do more than glance at it depends on how striking, interesting, and relevant it seems to be. To give an example, I used LinkedIn almost daily. I have seen the same ad multiple times in the last few weeks. Until I sat down to write this post, I had no idea which brand was being advertised or what the message was. I did know that I had seen the ad before and that my instinctive reaction was “irrelevant.” Part of the problem was the device that was used to gain attention. The image of the person goes from transparent to opaque and then the message appears, but by then my attention has moved on. Sorry, Accenture and Dr. Idemudia, but now that I have paid attention, Girls that Code might still not be for me, no matter how important.

2) The audience could not be bothered to figure it out

I have seen it suggested that if your message is complicated or hard to understand higher frequencies will help the audience realize what you have to say. Good luck with that. People are cognitive misers. They are not going to spend time trying to figure out what you are trying to tell them. They have better things to do. If you have not made your message simple and easily understood, frequency is not going to save the day (at least not cost effectively). I recognized that I had seen the Accenture ad before but there was nothing to trigger any interest to figure out why it might be relevant to me. Similarly, the person in a space suit that I have seen several times (no idea what brand that was for). And that frequent ad for IBM…but then, IBM, right? How interesting could that be?

3) The audience did not find the message relevant

I assume the IBM ad was targeted at me for some reason but targeting cannot make up for a lack of perceived irrelevance. The fundamental point is that you are advertising to a human being. Unless the person notices your ad and thinks it might be relevant to them, all your effort and money will be wasted. The immediate response to advertising depends on the ad’s ability to earn attention and deliver a clear, relevant, and motivating impression. If people respond to advertising – perhaps by filling in a lead-gen form – it is because they believe the ad is relevant to them then and there. And not everyone will be in-market now. Even if people have bought your brand before, checked out your website, or just have the right profile, there is no guarantee that they are ready to buy when they see your ad.

Is there a cumulative effect from frequency?

To some extent I believe it depends on the advertising objective. I have seen it suggested that multiple impressions in a short period of time somehow have a cumulative effect on behavior. The belief seems to be that people go through stages of Awareness, Interest, Action as frequency increases. Given the human mind takes only milliseconds to judge whether something is of relevance this seems highly unlikely. Even if the audience is a niche one, I see no point in building excessive frequency in the hope it will somehow beat people into submission.

If your intent is to drive immediate behavior, then you just need enough frequency to ensure that people notice and attend to your ad. People will buy you brand when they need it, not because they saw the same ad multiple times. However, if your objective is to influence future buyers with a brand building campaign, you need to build memories, and some frequency in a short time frame is a good thing. Repetition helps ensure that the impression is memorable, like learning a language. And if the advertising does not leave a memorable impression linked to the brand in people’s minds, it will not do its job.

How does recency factor into this?

Recency is a when an ad impression comes proximity to a purchase occasion, rather than frequency which is a measure of how often an ad is seen. For linear TV the recency rationale is that if new buyers are continuously entering the buying phase, it makes sense to spread your exposures out equally over time to maximize the chance that someone will be exposed close to making a purchase. So it is less about lots of impressions in a short period of time and more about spreading impressions out over a long period of time.

One could argue that in with digital media, particularly for more deliberative purchases, performance marketing follows a similar strategy, because you can identify when someone is researching a purchase. But for categories where there are no obvious digital signals of intent, or for brand building campaigns where people may not even buy your category yet, spreading exposures out across long periods of time does make sense. Not only does this strategy increase the probability someone will be exposed to your advertising close to purchase, but it will also help sustain impressions of the brand.

Where does frequency rank when it comes to effectiveness?

Back in 2017, Nielsen and NCSolutions (NCS) published a meta-analysis that examined the effects of 500 on sales of 863 campaigns for a variety of product categories that ran on both linear TV and digital platforms. Given that the time frame was 6 months and the dependent variable the incremental lift due to exposure to a campaign this study only looks at the short-term impact of advertising (yes, 6 months is short-term for many product categories). Intriguingly, the analysis does not even mention frequency, presumably because recency proved to be more important.

So, in rank order of importance by sales contribution we have,

1. Creative
2. Reach
3. Brand
4. Targeting
5. Recency
6. Context

It will probably come as a surprise to many that creative was even more important for digital than TV, contributing 56% to sales, and far more variable in its contribution. Why? The analysis finds that the quality of creative on digital is far more variable than on TV. Digital advertisers tend to produce a wide range of content and then direct spend toward the best performing ads. That is fine in theory, but the rationale falls flat when none of the ads are effective. Reach accounts for 22% of sales contribution, targeting 9% and recency 5%.

Similar findings for brand building

The Nielsen/NCS analysis looks at sales. By contrast, Kantar has published several studies looking at brand building campaigns (although the analysis is equally short-term because it is usually limited to the life of the campaign). In one analysis of the drivers of salience Kantar found that creative accounted for 50% of the contribution, reach 25%, and frequency 16% (the full webinar is behind this form).

Differences in measurement make it difficult to compare the two studies directly, but safe to say, both studies find creative and reach matter far more than frequency or recency. Given the different dependent variables, then we might assume that recency is more important to driving sales response and frequency more important for brand building (helping build and sustain impressions over time).

Questions to ask before worrying about frequency

1) Why will your target audience pay attention?

To paraphrase George Costanza from Seinfeld, it’s not about you, it is about them. No matter how important your brand or product pitch is to you, it is likely incidental to your target audience. Instead, figure out how will your brand or product make their lives easier, better, or more rewarding. Then, if you are using display or infeed ads, get that idea out front and center in simple, concise imagery and language. For video, particularly linear TV, you have a little more time to play with, but remember, video does not leave time for reflection. If they cannot understand what your video was trying to convey while watching, don’t expect them to think about it after the event.

2) What will make your ad and brand stand out?

No, the answer is not to force people to see the ad. That interstitial will likely just annoy them. Remember, it is not about you. How can you make your ad earn attention? For digital display, what striking but relevant image is most likely to trigger interest? (Test it. You are going to be a lousy judge because, remember, it’s about them.) Again, video is different. Online you have 5 seconds to make your pitch for attention. Once your ad has earned attention then you need to make a compelling impression, one that is intrinsically linked to the brand in people’s minds. Because, even if this is not a brand building campaign, and the recipient is not in-market right now, why waste the chance to make your brand more salient? Besides, maybe your brand is what people are most interested in.

3) How can I best reach my target audience?

OK, so you sell widgets, and not everyone buys widgets. But even if you have a niche audience why would you exclude potential buyers and stakeholders from your campaign? Particularly for brand building campaigns your objective should be to widen the sales funnel, and that implies reaching as many people as possible for your budget. All the evidence I see still suggests that TV is still the best reach building media channel and digital struggles to come close. But then, that is a function of the reach of the digital platforms themselves, what you can buy on those platforms, and the mindset of the person doing the buying. If buyers prioritize frequency over reach, what do you get? Low reach.

Done all that? Now you can worry about frequency. Or recency. Maybe you should test what works best for your brand and its campaigns? But remember, you must interpret the data. If it takes lots of impressions to produce the optimal response, you advertising is probably a dead dog.