RORY SUTHERLAND: What financial services can learn from cafés and comedians

Rory Sutherland is Vice Chairman at Ogilvy, as well as an author and marketing expert.

What makes a comedian funny? Yes, you could write a whole thesis about the nature of humour, but, looked at backwards through the lens of Darwinism, this question is not as hard as it sounds. It’s simply a question of natural selection. Comedians tend to be funny because they receive instantaneous, minute-by-minute feedback on their set– usually in the form of laughter, or the opposite. If you try your hand at stand-up and people don’t laugh or, worse, throw things at you, you may make a few more attempts, but after that you’ll probably quit. Hence the few people who survive in comedy tend to be quite funny by definition: the also-rans have been rapidly weeded out – either by withdrawing from the game or though not being invited back.

Successful comedians are also committed Darwinists in their approach to their own work. When they first break new material, they will continue to tweak the order, timing and precise phraseology of jokes to try to maximise the laughter. It’s why comedy tours tend to start in smaller, provincial venues before progressing to the big stage. By the time you see a routine on television, it has been through countless slightly different iterations before the artist settles on the final content.

This sometimes has the effect of making comedy look easy. Survivorship bias means that the comedians we see are overwhelmingly the successful ones, and we are most likely to see their routine at a late point in its development where it is at or near perfection. Hidden from our view are the many dud comedians and gags which fail to land, which were weeded out early in the game.

It’s similar to the comment “Didn’t the Elizabethans build houses well – they are still standing over four hundred years later. True, but the Elizabethans also built a lot of rubbish houses. They just aren’t here for us to critique.

The same applies to restaurants. It is very easy when walking down a street packed with packed eateries to conclude that the hospitality industry is a licence to print money. But, as Nassim Taleb remarked, “The graveyard of failed restaurants is a very silent place.” The odds are that, on the site of every booming bar or café, four or five attempts previously failed. In my home town, there was a restaurant location where, one after another, countless brave people failed even to draw a small clientele. When the local kebab shop owner finally bought the place I was in despair: “God, please don’t do it!” He opened an upmarket Turkish restaurant, which is an astounding success. Within days it was obvious that he was right and I was wrong – he had found the motherlode.

The value of fast feedback is easy to overlook. It’s why geneticists perform so much research on Drosophila melanogaster, a species of fruit fly which is used as a model organism for experimentation on account of its simple genetics (it has only four chromosomes) and its extremely rapid lifecycle. If you tried performing genetic experiments on elephants, you’d be long dead before your work attracted the attention of the Nobel committee.

My contention is that a great deal of optimisation work in marketing focusses on those parts of the marketing mix where feedback is fast. Most notably acquisition and performance marketing. This is not unimportant: it mostly makes sense to optimise the funnel from the bottom upwards, since there is no point in endlessly feeding a funnel with a very narrow spout. But this does mean that far too little time and money are spent on longer term activities such as customer retention – not because they are unimportant, or because they are immeasureable, but because measurement of these effects is very slow. Imagine being a comedian with a very slow, humourless  audience who laugh five minutes after every joke. Not easy.

This problem is particularly acute in categories such as financial services, where customer retention tends to be fairly long in any case, making measurement interminably slow. What we perhaps need here is the use of what biologists would call a model organism: a fruit fly, in other words.

When I give speeches, I sometimes tell jokes, but I cannot use laughter alone as a feedback mechanism – believe it or not, sometimes I like to make a serious point. So I use a proxy for audience engagement: quite simply, if the AV people are paying attention to your talk, you’re doing okay. The paying audience have to pay attention for reasons of politeness and convention: the AV people only listen when you are interesting.

In the same way, it might make sense for financial institutions not to attempt to prove the value of retention activities in their own environment. It takes too long. Instead, why not pay fast-feedback businesses to perform the experiments for you? Online supermarkets and restaurants, to give just two examples, can perform experiments on repeat business and get results in weeks rather than years. I am a very strong believer in the value of giving people branded merchandise – and recently I learned of an experiment where an online supermarket had given away a small gift and enjoyed a monumental and immediate effect on customer retention. McDonald’s achieved spectacular success by simply giving children balloons – which even increased sales among adults.Banks can’t try these things – but they can learn from them.

No, I agree, a restaurant is not a bank, and the evidence is not wholly conclusive. But it’s still better than doing nothing at all, which is often the only alternative.