Advertising today causes sales, some next month, most much later, on and on. 

Most of the advertising that underpins this month’s sales aired many years ago.

The purpose of advertising is to encourage sales, i.e. to make it more likely that more consumers will buy your brand on more of their purchase occasions. There are 5 primary ways advertising does its job (these are described in Chapter 9 of How Brands Grow), one of which is the most common mechanism (i.e. enhancing mental availability).

This short article is about something different. It’s about when the sales impact occurs, when any effect can be vaguely seen, and when it likely can’t. And in turn, what this means for setting advertising budgets. 

It’s often thought, that if advertising is successful in encouraging sales then this should be able to be seen in a brand’s sales figures, and likewise if it fails to show in this month’s sales then it can’t be working. Of course, advertisers are aware that many other things push sales up and down, e.g. weather, competitor advertising, pricing, in-store display, supply issues, and so on. But it is felt/hoped that “econometric” statistical analysis can control for all of this and tease out the true causal impact of advertising in exact dollars. Unfortunately this is overly optimistic, statistical analyses can help clear up some of the noise but it can’t reliably quantify the overall sales effect.

Some of the things we do in marketing are to catch our fair share (or better!) of this month’s category buyers. We do this by improving the physical/purchase availability of our brand or the brand’s cost to the buyer this week. We use many levers, such as paid search & display (e.g. Google, Amazon), in-store display, price specials, free delivery offers, and so on. Along with our online and bricks & mortar listings, opening hours, (lack of) stock-outs, available sizes/colours/flavours. These levers affect how many of this month’s category buyers buy our brand compared to our competitors. Which means that the sales impact of these tactics can be seen immediately, and in full, when we turn them on, and when we turn them off.

This means we can (and should) work out how much to spend on these “catch them as they fall” activities, ie which work best, and the point where spending more yields little return.  Working this out is important, good management.  But it has little to do with how much we should spend on advertising, i.e. there is no ideal 60/40 ratio (see Report 107 Activation vs Advertising).

Broad reach advertising’s impact on sales is very different. Partly, because most of the people that this month’s advertising hits are not going to buy this month, or even within the next year. But mostly because advertising builds/retains mental availability and this affects many purchases beyond this month (sometimes for a lifetime).

Of course some of our advertising exposures hit people who are about to buy from the category, but not many. Even most “fast moving consumer goods” are bought fewer than ten times a year on average, i.e. only in fewer than 10 weeks out of 52. But this is a highly skewed average, the vast bulk of category buyers buy only once or twice a year, so the vast bulk of advertising exposures will have no influence on this month’s sales. For many services, B2B, durables and luxury goods most buyers won’t buy the category this year or next. Our advertising today that reaches these buyers can only affect far future sales. Similarly, the advertising that underpins this year’s sales aired many many years ago.

This is why our broad reach advertising is very different from the things we do to catch our share (or better) of this week’s category buyers (see Report 107 Activation vs Advertising). And must be evaluated differently.

What about when we do see a sales uplift?

Sometimes the effects of a new campaign are discernible in this month’s sales. Comparing our sales to a seasonal trend line will help see if the effect is real.

What does it mean if we see an effect? Well, sometimes it’s just that we got lucky, e.g. it aired when there was little else going on in the market, and so we need to be a bit careful about concluding that this is fantastic creative and/or media strategy. But it’s certainly a good sign that the creative content of this advertising is working (corroborating evidence can be collected from in-market surveys, e.g. do people who have seen the ad register the brand correctly). In which case this is advertising worth investing in because its long-term effect will be even greater.

Young Brands

It’s easier to see advertising effects for a new brand. More consumers are becoming aware of the brand’s existence and, partly because of this more retailers are listing it. The aim of advertising for young brands is to establish a foothold of (overlapping) mental availability and purchase availability. To give the brand a chance of growing.

An important but overlooked role of advertising is to tell new buyers of the brand what was the brand they just bought. Hitting a consumer with an exposure after they purchased (and therefore long before most will purchase the category again) seems like a waste of money (no ROI/ROAs!), but it helps to lay down memory structures that are essential for repeat-purchase. All of us have gone to a new restaurant, bought a new yoghurt, visited a web site… and then never repeat-purchased – not because of dissatisfaction, not because of lack of need, but simply because we never properly laid down memory structures for that brand, or at least that its mental accessibility remains very low. Advertising is important reinforcement. 

Hopefully a new brand’s sales are rising and they tend to rise faster if more is spent on advertising. Advertising is oiling the wheels of innovation, and keeping existing brands on their toes. Even economists approve of advertising in this context.

So for small brand it’s more likely that some of the sales impact can be seen in this month’s sales figures, but remember most of the effect is a long-lasting one on mental availability and so can’t be quantified correctly by looking that this month’s sales increase.  Advertising is helping to build an asset that can last decades.

Be wary of the tendency to burst spend on small brands just because sales bumps can be seen.

Established Brands

Once a brand has substantial availability (mental and physical), then the effects of advertising are hard to see, both on sales and on memory metrics. So how can advertising budget be set, and how can this spend be evaluated?

Setting ad budgets

Spending on purchase availability can be judged by its impact on this week’s sales, because the full effect occurs this week due, and so this spend can be optimised. The aim is to spend to the point of declining returns.

But advertising exposures (a) mostly reach buyers who won’t buy this week, and (b) will build/maintain mental availability that will affect many future purchases. So how much to spend is a forecasting problem, i.e. forecasting the potential future of the brand. Contrary to popular belief over investing in advertising will in no way guarantee a brand growth, but advertising will help if the brand is also gaining purchase availability, if its portfolio is improving and growing, if demographics and preferences are moving in its favour. These things need to be forecast.

Brands with an unfavourable future can not be saved by advertising spend. Brands that are destined to hold their share need maintenance advertising budget, but no more.

The Institute has a sponsor seminar and specialist research service to set evidence-based advertising budgets based on the principles of forecasting science.

Evaluating advertising

When sales go up a lot of people claim it was due to their work…. sales team, R&D, ad agency, media agency. But as this report has explained, odds are sales won’t change their trajectory when adjustments are made to creative or to media strategy. So how can we tell if we are doing a better job at advertising?

Single source data capturing both advertising exposure and purchasing (see Report 36 Individual-level Advertising Sales Effects) for individual buyers can be used to evaluate the sales power of different creative. Such data will become increasingly available but it is still rare. So we recommend immediate in-market post-tests to evaluate creative impact.

To evaluate media strategy we recommend avoiding methods that depend on weak correlations to sales movements (e.g. marketing mix modelling), and instead using media metrics to assess spend against evidence-based principles, i.e. to improve branded reach, to reduce bunched exposures, to increase working media. We recommend appropriate metrics that can show the CFO that media spend is improving, in the same way that appropriate metrics are used by production to show that factory spend is improving, and indeed that the finance department use to justify their work (‘days sales outstanding’ for creditors, amount of free cash for cash-flow management, and so on).

Good advertising may, but probably won’t, show its sales effect in this month’s sales figures, but advertising can still be accountable.