Andrew Willshire: Focus on the short term and the long term will take care of itself

One of the more regular events in marketing is the high priests descending from their mountain to condemn those humble sinners who have succumbed to the illicit temptations of short-term thinking. Like bible-thumping preachers, they warn of hellfire and damnation if we don’t change our ways. The trouble is that I’m not sure it’s making advertising better.

And so, like a pound-shop Martin Luther, I step forward and nail to media’s cathedral door, not 95 theses, but just one: Create a strong short-term effect and the long-term will take care of itself. Before digging any deeper, let me state some caveats:

• The product or service needs to be good. As Bill Bernbach said, nothing kills a bad product quicker than good advertising.
• The advertising must not be tied into a price promotion that will undermine brand value in the long term. Any fool can sell a product by slashing the price.
• The measurement must be robust. I’ve discussed digital attribution in depth here; suffice to say, it’s not the droids I’m looking for. Good econometrics will provide excellent estimates of short-term sales uplifts. Measuring long-term effects via econometrics is difficult and imprecise due to the diffuse nature of repeat purchases and the lack of an independent reference with which to calibrate estimates. For measuring the long-term impact, loyalty-card data, which enables tracking of individuals’ repeat purchases, is superior.
• I’m talking about real metrics. Click-through rates and brand awareness indices doth butter no parsnips; profit is the only metric worth a damn. And provided that your price remains constant, sales volume is an excellent proxy for profit. That’s right – marketing is about selling. Who knew?

Despite the distinct sound of pitchforks being sharpened in the distance, I will press on. The key question to the advocates of focusing on the long-term is this: how can selling a good product in the short-term possibly harm the long-term value of a brand?

READ MORE: How to balance short-term impact with long-term results

However much we love our campaigns, no amount of advertising beats the product being in the living room, on the kitchen shelf or sitting in the driveway – a constant reminder of the brand and of a promise fulfilled.

As such, the best predictor of which brand someone will buy in the future is the brand they buy already – it’s the most ‘mentally available’. In the car market, repeat purchase rates can be as high as 70%. This isn’t just down to some abstract notion of brand love either. As Byron Sharp put it: “Loyalty is often more a function of habit, familiarity and lack of caring rather than unbound devotion.” Marketers should be aiming to create that habit and familiarity.

Most measures of brand awareness are also strongly correlated with market share, regardless of the amount of advertising. The combination of mental availability and post-purchase rationalisation means that measuring brand awareness will tend to reflect recent purchases rather than drive them.

During the Korean War, the US Air Force strategist John Boyd created the concept of the ‘OODA loop’, referring to the decision cycle of observe, orient, decide, and act. Originally related to aerial combat, it was later broadened to all manner of competition. The key insight was to get inside your opponent’s decision-making process, such that they didn’t have time to figure out what you were doing and effectively respond. This is almost a textbook description of marketing – short-circuit the customer’s decision-making process and make your brand the obvious choice when they pull the trigger on a purchase.

Reach consumers when they’re ready to buy

Crucially, though, getting inside the loop means reaching consumers close to the point of purchase. It follows that the most effective marketing strategy is to convert the people who are going to buy tomorrow, not next month and certainly not next year. If you reach them too early, you leave them open to your competitors’ messages, blunting the impact.

This isn’t a new idea. Erwin Ephron wrote about recency planning back in 1998: “Recency planning ignores purchase cycle, because it targets the purchase not the consumer who makes the purchase. As long as there are purchases each week, it doesn’t matter how often, or seldom, the average user buys.

“Whether a consumer is ‘ready to buy’ is more important than the number of messages the consumer receives. When a consumer is in the market, a single message can have an effect, but if a consumer is not in the market, multiple messages are not likely to make the sale. Reaching three consumers, once, will generate more purchases than reaching one consumer three times.”

Awareness is most useful close to the point of purchase and the best driver of sustained brand awareness is consumers buying the product

There is some evidence to support this theory. Some years ago, some colleagues and I were contracted to build econometric models for a large car manufacturer. This was the first time that we had modelled this sector and we discussed what we expected to see. Having read the literature, we ‘understood’ that a car purchase was a major decision, made over a long period and influenced by thousands of adverts. As a result, we thought that identifying short-term sales uplifts would be difficult. We were wrong – they leapt out of the models.

In fact, the advertising was reaching people who were on the cusp of making a purchase and influenced their actions by, for example, inclining them to visit that brand’s dealership first. Long-term effects, on the other hand, were impossible to isolate. And no wonder: some studies estimate that the average person sees around 50 TV adverts every day. Given that most researchers think the total effect of advertising is relatively weak, why do we give credence to the idea that an advert from six months ago is a significant as the one seen yesterday?

The focus on the long term has led media to be planned differently. It is common to define long-term brand-building media as having a broad reach and being highly emotive while short-term media is tightly targeted and information-rich. The trouble is that this artificial division splits apart two of the vital roles of media – broad reach and persuasion – leaving us with advertising that fulfils neither role adequately. What is wrong with communicating the benefits of the product to a large audience?

Instead we have two types of underperforming media: intrusive, minuscule-reach, high-frequency digital advertising, which companies like Procter & Gamble are abandoning due to the total lack of impact; and big showpiece campaigns with high reach but content that is only vaguely related to the actual brand (see Mark Ritson on the recent Heineken advert). When neither type of advertising does much to drive sales, we have a problem.

Long-term thinking also provides an excuse for poorly performing media: ‘There’s no short-term impact but don’t worry, it’s contributing to long-term brand awareness – trust me.’ Note that this leaves no space for identifying media that just isn’t working. I would contend that if media has a long-term effect, then it must have a short-term effect. Media that doesn’t change behaviour tomorrow will not change behaviour six months from now.

To clarify, I’m not saying that brand awareness isn’t important. I’m saying that awareness is most useful when it receives a boost close to the point of purchase and that the best driver of sustained brand awareness is consumers buying the product, using it and liking it. This means that the most important factor is to drive an individual’s initial purchase and to target as many purchase moments as possible.

So let’s take a break from fretting about the long-term. In another context, John Maynard Keynes said that “in the long run, we are all dead”. Ephron said much the same for marketing: “If you don’t get enough next purchases, building a brand doesn’t matter.”

Andrew Willshire is the founder of analytics company Diametrical and former global director of advanced analytics at Maxus.