You should not buy additional exposures close in time (eg same night, day, week).

How close is close? That depends on the size of your budget, if it is very large then spreading out your advertising means aiming for most consumers to see your advertising every week. For most brands with smaller budgets it will be something like every month or quarter.

This best practice (see Ehrenberg-Bass Institute Report #66) is far from actual practice. Today, many brands still blow their budget in a short period where some people see their ads multiple times in a single evening, and then we all go weeks or months without seeing any.

The principle of spreading out advertising exposures is supported by:

• laboratory experiments
• learning theory
• memory theory
• real world single-source data
• academic modelling of sales effects of different patterns of ad spend.

Different researchers, different methods, same finding.

A second exposure, close to the first, has less impact than the first. And a third has even less than the second. Single-source experiments suggest that for grocery brands even a second exposure within a few weeks can have less impact than the first exposure.

Similarly, a 60 second advertisement does not have double the effect of a 30 second advertisement, which in turn does not have double the effect of a 15 second advertisement.

But not all the exposures I pay for actually occur? Shouldn’t I compensate for this?

Unfortunately viewers/listeners/readers can be quite distracted, this has long been the case. There are other things on and off the screen for them to look at. So even without active ad avoidance (eg ad blockers, leaving the room during an ad break) many “opportunities to see” (OTS) don’t translate into real exposures. And many exposures are partial, which is the simple reason why longer advertisements work a bit better, or why adding any audio branding to visual branding improves brand recall (Ehrenberg-Bass Institute Report #103).

Now, in a movie theatre where near 100% of pre-movie OTS are seen, paying for a 2nd OTS will give a frequency of two for practically everyone in the theatre. And this would be a rather pointless media buy, as the second exposure has far less impact on those consumers’ memories, and yet costs much the same.

But what about a medium where only 50% of OTS turn into real exposures? Should we therefore buy more close OTS to make up for the lack of noticing (distracted viewing)? Well here’s what we get in this situation if we buy a second OTS during the same program, web session, magazine etc… Say our first OTS reached 100 people of whom 50 watched the advertisement, our second OTS turns that into:
– 25 people having two real exposures (of which the 2nd has less impact)
– 50 people who have had one real exposure
– and 25 people who have had none.

We’ve doubled our ad spend and yet achieved not much more than 50% increase in effect. If we buy a third OTS, the outcome is even worse. Alternatively, we could spend our money on a second OTS on another evening, or on a different channel/website. The effect on reach would be greater, and hence a greater effect on memory/sales.

This evidence leads to one lesson: don’t bunch your ad exposures.

Examples of (wasteful) bunching:

• program sponsorship
• a row of billboards on the same street
• multiple ads in the same publication
• failing to “frequency cap” on social media