Every marketing manager’s dilemma: should the marketing budget be spent on media advertising, with the aim of building brand image, or on promotion, with the aim of getting the brand moving on and off the shelf? There must be a point of balance: too much emphasis on promotion results in disloyal and fickle consumers, too much emphasis on advertising equally can give you the most-loved but least-bought brand.
How do you find the right balance?
Your ad agency (and let’s face it, the glamour seeker in you) pushes you towards glossy ads. The retail chains demand ever-increasing promotion. If you try to be scientific, by looking at the results of adopting different mixes in different regions, you get measurable results on sales growth (usually highest in heavily promoted regions) but few or no results on brand building (so no way of justifying more advertising). Yet if you are in a FMCG business for the long haul, the greatest asset you can have is a preferred and trusted brand.
Previous PIMS research has shown (Fig. 1):
- the drivers of promotional spend are different from advertising
- high relative advertising builds customer preferences
- the ad/pro mix is correlated with competitive strength
- increased advertising and promotion boost market share
A business’s potential economic value
Figure 1: Impact of advertising and promotion spendings
Another piece of PIMS research, published in the Harvard Business Review, spells out the importance of another key balance for consumer products, between relative advertising and relative price versus competitors.
Other researchers have also found that advertising builds brand strength, and is helpful in shaping consumers’ expectations. In addition to the PIMS evidence, there is considerable further research demonstrating the trend over time away from advertising towards promotion.
Figure 2: Optimum ad/pro balance
This PIMS Letter demonstrates two remarkable new findings (Fig. 2):
- The optimum ad/pro balance depends on a few key characteristics of the brand
- The optimum ad/pro balance does not depend on the total marketing spend, either in absolute terms or relative to competitors
The PIMS Evidence
Over the years, PIMS consultants have worked with thousands of groups of client managers to assemble key information on their businesses, covering production, finance, customers, distribution, competition etc. This particular study uses 300 of these businesses that are FMCG brands, including many familiar household names, roughly half each in Europe and North America.
The success measure we use is four-year average return on investment (ROI). This measure is preferred to measures of sales volume usually favoured by marketing analysts, because there is no point in increasing sales if you don’t make money doing it. Conversely ROI is preferable to looking at return on sales (ROS), because there is no point in increasing margins if as a result you lose sales fast. Four-year averages smooth out the business cycle and the day-to-day cut and thrust of competition.
When we look in this database at the effects of advertising and promotion, on average they are much as you might expect: good profitability depends on the mix of advertising and promotion, whereas gaining market share depends on the sum of the two. This paper is concerned with investigating the first question – getting the balance right – which is usually an issue tackled within the marketing function. The second question – the size of the overall marketing cake – is a much wider strategic question involving all functions of the business, and cannot be addressed in isolation from the full range of strategic factors (market attractiveness, competitive strength, value added structure, human resources).
Figure 3: Average profitability resulting from different mixes and brand ranks
So what is the best “ad/pro” balance? Is it 75/25, or 50/50, or 25/75? The PIMS evidence shows that the answer differs between brands depending on seven specific factors. First look at market share rank. In the chart (Fig. 3) we have shown the average profitability resulting from different mixes and brand ranks, circling the highest number in each row.
“So what is the best ‘ad/pro’ balance?”
If we look in the bottom row, we can see that life is usually tough for number 4 brands or smaller. The best they can do is earn an average 11 % ROI, probably less than their cost of capital. The best ad/pro balance seems to be in the region of 40 % advertising and 60 % promotion, with a steep fall off in profitability as you move away. Worst of all are the low share businesses with high advertising, who on average lose money in their attempt to play on the big boys’ turf.
It is no surprise to business managers that brand leaders are, on average, more profitable. The interesting fact is that they can generally afford to play the high-stakes advertising game and reduce promotions to a quarter of the cake or less. Ad-vertising is particularly scale sensitive: if you have double the sales of a competitor, and spend only the same percentage of sales, you get more than twice the impact. You can get the best creative brains, the best discounts, and savings on advertisement production costs.
However, the table (Fig. 3) also shows that the profit penalty of moving away from the peak is relatively small: even with 40 % advertising and 60 % promotion the average ROI for lead-ers is still 39 %. In practice, this is what many market leaders are increasingly doing. Another split of the database can help to understand why (Fig. 4): here the rows represent different levels of channel concentration.
Figure 4: Different levels of channel penetration
Again we have circled the highest number in each row. The table contrasts sectors (bottom row) with highly concentrated channels against sectors (top row) where there are still many fragmented outlets. Many brands are seeing their channels get more concentrated over time; marketing managers face fewer and more powerful customers pressuring them to promote heavily – to keep the brand moving through the channel. Again, business managers will not be surprised to see powerful customers depressing profitability – but they may well be surprised that customer power is no more important than whether they get their marketing mix right or wrong. Even with a few customers, it is vital to keep a reasonable proportion (over 30 %) of advertising in the mix.
Figure 5: Levels of real market growth
Figure 6: Corporate versus brand advertising
So far, though, the findings cannot come as a big surprise. The next finding is powerful, unequivocal, and at first sight often counter-intuitive. We can split the database between different rates of market growth (Fig. 5). Circling the highest number in each row shows a strong diagonal pattern: you should advertise in declining markets, promote in growth markets. Note too that, whatever you do, it is difficult to earn more than 20 % ROI returns in a rapid growth FMCG market. The underlying explanation seems to be that competition is very different in a boom. Dozens of entrants jostle for leadership, and brand franchises don’t count for much with inexperienced consumers. Marketers must be single-minded
Figure 7: Profile chart
about getting their product on the shelf and stimulating con-sumer trial: they do not need to advertise. Advertising can wait until the market is bigger and consumers more mature. At the other extreme, there is no point in pushing on a string by promoting in a declining market: the survivors will be the well advertised brands that keep their image up to date, retaining consumers who may be thinking of giving up.
“What about corporate versus brand advertising?”
What about corporate versus brand advertising (Fig. 6)? Some companies, like Heinz and Nestlé, get mileage because every advertisement for one brand actually advertises its sister brands too. This scale advantage is much more achievable with advertising than with promotion, so when it can be achieved it releases resources that swing the optimum mix back a bit towards promotion. Many marketing managers reading this will feel like a yo-yo. First the evidence tells you to push advertising up, then down, then up, then down again. If you want to complete the confusion, try plotting your business on this “profile chart” (Fig. 7), which contains three additional factors that affect the optimum balance:
Again, the success of promotion-driven innovation may be a surprise, but the trend has been widely noted.
You will almost certainly have some characteristics pointing to promotion, and others to advertising. PIMS researchers have developed a mathematical model that works out the optimum balance for your specific business. By looking directly at the experiences of analogous businesses in our database, PIMS also calibrates the dynamic consequences (share gain etc.) of alternative total spend as a percentage of sales, allowing an optimum allocation across a portfolio of brands.
Figure 8: Factors not affecting optimum mix
The Dogs That Didn’t Bark
As Sherlock Holmes found, sometimes it is the things that don’t happen that are the most instructive. For this piece of work, we looked at many other factors beyond the seven that we have reported on (Fig. 8).
The three findings in the left hand column underline a common PIMS finding; that what matters is your structural stra-tegic profile. The fact that total spend, and spend relative to competitors, do not affect the optimum balance point is a confirmation that while advertising and promotion have their separate cost/benefit equations, the relative impact of each is fairly constant in a given situation.
Limitations To The Research
One concern we couldn’t fully cover was the position in products (e.g. tobacco), or countries (e.g. Sweden), where TV advertising is banned. The argument could go either way: if advertisements are less cost-effective, you could argue that the mix should switch to promotion to maintain cost-effectiveness. Alternatively, if success is a matter of getting the right balance of brand building and consumer trial, marketers should redouble their efforts to present the brand favourably to consumers via magazine advertising, posters, sponsorships etc.
“It is never the job of the marketing department to stimulate consumer disloyalty by promoting without advertising.”
The second argument is more in tune with the research so far: it is never the job of the marketing department to stimulate consumer disloyalty by promoting without advertising. Another concern might be the direction of influence: correlation does not prove causality. But to argue a reverse causality requires bizarre speculation. You would have to agree that successful number four brands say that “because we’re suc-cessful we’ll spend 40 % on advertising”, whereas unsuccessful ones say “because we’re unsuccessful let’s either spend it all on advertising or all on promotion”. You would further have to argue that as you move up to the market leader the selected mix of the successful players shifts smoothly to 60 % and then 80 % on advertising and that suddenly no unsuccessful number ones plump for all advertising.
We can see no possible behavioural or economic logic for such hypotheses. It is of course possible that each cell of every matrix has a unique underlying factor that causes a business both to be in that cell and to have the given level of ROI. We invite testable suggestions along these lines.
The final caveat is that our sample was fast-moving consumer brands: not private label, not consumer durables, not business- to-business. This sample has great homogeneity in channel structure and few complexities of technical service, warranty claims, etc. We hope in future to be able to identify other large samples in the PIMS database with sufficient homoge-neity to make prescriptive statements about marketing. In the meantime PIMS consultants address specific cases via the analysis of smaller samples of “look-alikes”.
To use this research in practice, we do not recommend marketing managers to follow slavishly the output of the PIMS model. As consultants, we follow an interactive approach where the complex specifics of your situation are cross-ref-erenced against the evidence of our database. The resulting synthesis is then both workable and objectively validated.
Conclusions
The marketing function can make a big difference to the bottom line by getting the balance of it spend right. When we look in our database at how businesses actually behave, it appears that many brands – particularly market leaders and those in static/declining markets – do not devote enough resource to building and refreshing their image with consumers. This is partly due to the short-term pressures from the trade, and partly due to the short-term bias derived from measuring marketing campaigns by their effect on monthly sales volumes rather than underlying consumer preference. This research should help many more get it right.
