Short formats are a brilliant addition to our creative arsenal, but the discussion around them is promoting a race to average, writes Thomas Wagner, Planning Director at BBH Singapore.

There is a convincing body of research showing that on average 15 second TV ads are as effective as 30 second TV ads on a number of metrics, such as advertising recall, likability and brand identification. Quoting from a 2010 Ehrenberg Bass study by Kate Newstead and Jenni Romaniuk in the Journal of Advertising Research:

Although half the length of a 30s ad, the media cost of a 15s advertisement is between 60 to 80 percent that of a 30s advertisement. Is this price premium warranted on the basis of relative effectiveness? We find that recall and likeability scores for 15s advertisements are approximately 80 percent that of 30s advertisements. However, the performance of 15s and 30s advertisements was equal for correct brand identification. The better branding execution of 15s advertisements appears to compensate for their length disadvantage. This difference in branding execution is consistent across advertisements sourced from Australia, the United States, and the United Kingdom. These findings will aid marketers to commission effective advertising and provide useful benchmarks for negotiating deals for media purchasing.

Not only did the 30s ads perform just marginally better on recall and likability, they do worse when it comes to correct branding score.

Michael Sankey and Ken Roberts in their March Admap article use a different methodology but get to the same conclusion:

The sober minds at the Ehrenberg Bass institute consequently conclude that if a marketer wants recall, likability, branding and efficiency – and I have yet to meet one who doesn’t – a 15s ad is a fair buy at 80% of the price of a 30s ad. Or if branding is the measure of choice, 15s ads are an even better deal.

Fast forward eight years from the Ehrenberg Study, 6 second ads are the talk of town and the people at MediaScience and Ehrenberg-Bass have run another study.

Short-form advertising often punches above its modest duration when it comes to effectiveness. As an example, the two organizations found that the average unaided brand recall from a 7s ad stood at 9%. That total, by contrast, reached 15% for traditional 30s messages and 12% for 15s communications. [ … ] In fact, if everything was kind of proportional [in line with ad duration], you would have seen a 3% effect. What did we get? We actually get a 9% effect.

It turns out things are not proportional. All seconds are not created equal, the vast majority of the impact seems to come from just a handful of them. Traditionally media schedules had a 30s version that attempts to wrap a more or less ‘engaging’ ‘story’ around the promise and a 15s version that just focuses on the promise. Today, you can make even more efficient, better branded 6s ads, and buy attention and brand recall at reassuringly predictable rates. Which is why Facebook are testing ad-breaks and YouTube are throwing their marketing weight behind 6s non-skippable bumpers. While they are rightfully doing their best to convince the industry that these assets are not just great for media efficiency and forced attention, that’s exactly what makes them such a great sell, and the ‘Rookie of The Year’ of Ad Formats in this years’ media plans.

It makes sense. You would do it. I would do it. Every marketer with a CFO over his shoulder would do it. The predictability is alluring.

And yet, it sends us in the wrong direction.

Because the research would also invite an alternative reading. The clue is in the word average. Average liking of the ads in the Ehrenberg Study is 3.2 on a 5 point scale. The statistical equivalent of ‘meh’. Average brand identification is 40%. That’s 6 in 10 people not able to identify who was trying to sell them in the first place. Average recall is 50%.

On average, advertising is average.

Averagely effective with the right average media weight behind it but incapable of outperforming what the average competitors with the same average ad and average media spend do.

These findings shouldn’t be a surprise to anybody who has ever watched TV or more recently online video advertising. Much follows a tried and tested, average formula that has been internalised in most large advertising companies as what actually gets media investment – never mind the odd purpose film. An average ‘insight’ – or rather a ‘need’ – that sets up the need for the product, followed by the product, a reason to believe and ‘emotional’ pay-off, maybe with an average of 2 or 3 messages thrown in for good measure.  The 15 seconders at least leave less room for misunderstanding from cramming too many messages in. They have more branding per second.

It is only because the vast majority of 30s of advertising is 1) not actually engaging and 2) typically horribly branded, that 6 or 15 seconds of it do as much good as 30s. Even the Ehrenberg scientists seem to think that if advertisers made better 30s, at least from a branding perspective, the results wouldn’t be so damning:

If marketers are using 15s advertisements instead of 30s advertisements to improve marketing efficiency, our study indicates that such a move does not result in a major compromise in terms of effectiveness. (Fabian, 1986). However, branding execution analysis highlights that there is room for improvement within 30-second advertisements. If the branding execution were improved in 30s advertisements, they might then be relatively more effective than 15s advertisements in terms of correct brand identification.

Sadly, for all the reasons outlined by Lucian and Tom non-average does not happen very often, or very consistently, or, to be self-critical of the advertising industry, very well branded.

So if you are a marketer and only ever had the intention, guts or capability to make an averagely forgettable, averagely badly branded 30s ad, you can now improve your efficiency by turning it into a better branded, forgettable 15s ad, even a 6s one, whilst working less hard and making the CFO happy. This allure of a safe but small improvement over a slightly more uncertain but vastly bigger improvement seems to lead many to plan exclusively for 15s and 6s, and not even attempt to – for example – generate fame, the type of media money can’t buy. This optimisation is threatening to reduce much of the canvas for what could potentially be non-average.

Now, this is not about bashing short-form assets. They are a great opportunity to build clever campaigns. It’s an appeal against average. When research says on average short is as effective as long, this neither means that 30s cannot produce better results – they obviously can – nor that the short advertising we are now forcing people to watch should be the same averagely forgettable work that we subject people to when we think we have their attention, just with better branding per second.

It means an opportunity for those marketers who want beyond average returns to stand out with non-average media plans around non-average creative work. As Keith Weed only recently reminded the industry – and perhaps his own people:

The biggest return on investment is a great bit of creative. We spend a lot of time talking about data-led marketing and the ability to target better, and you need to do all that, but a great bit of creative will outstrip anything you do on media optimization. If you have a great bit of creative you still want to optimise media, but I think in a data-led world there is a lot of focus on that and we’re not focusing enough on the power of great creativity.