Digital advertising has a statistics fetish. It needs to stop.

As technology’s role in advertising has grown, allowing ever more advanced and hyper-targeted campaigns — to the point of “sniper targeting”/extremely elaborate pranks (which we can confirm are worth the money), and ever-more segmented targeting. The options are near endless. Now, more efficiently than ever before, your campaign for married men and women who like Pina Coladas and getting caught in the rain will find the right people.

Recently, I came across this harrowing yet brilliant article, about just how messed up “Escape (Pina Colada Song)” is. Essentially, it’s a massively broken relationship that doesn’t work and nobody is happy. A bit like digital advertising, as it happens.

A lot of the problems with digital advertising stem from a mix of poorly set goals, and poorly managed expectation.

Leaning on the song for a while longer, imagine you’re the owner of the bar called O’Malleys, the bar where troubled couples meet. When O’Malleys launches their digital campaign to attract those want away partners writing personals in the newspaper, they would probably work on simple assumptions:

1. People will see my ad.
2. The ad will show that our bar is for them.
3. These people will visit my bar.

In reality, that’s not how it works. Even in the song, it’s still not how it works. The audience targeting is miles off. For a start, something like this would be better:

1. They have been in a relationship for a few years.

2. They dislike yoga.

3. They like the dunes and the cape. 

4. Are known to be night owls.

5. Are recent advertisers.

6. Read the local newspaper

7. Exhibit repetitive check-ins and routines in the local area.

8. Like champagne and Pina Coladas

9. Live near O’Malley’s bar.

But even then, that requires the desire to escape from a humdrum relationship to be in place, the nerve to go through with it, and for them to want to choose your bar. The song may have pre-dated Ashley Madison, but this is 2018, and there’s a lot of work that needs done if you want attention. That work starts long before the desire for escape is even there.

O’Malley’s should be known as the escape pod — the place that couples don’t want to use, but if they need to, they know it’s there.

The trouble is, constantly raising awareness of the marriage escape pod bar is really bad for conversion metrics (and PR, one would imagine). Especially if we measure success based on the number of unhappily romanced individuals who visit in a given date range. Presumably in this example of orchestrated cheating, the bar does a survey or estimate based on how many tan lines in lieu of wedding rings are observed on the ring fingers of visitors (I’ll leave this to someone better informed on primary research methodology). But measuring success based on a simple metric alone is almost certainly the most dangerous way to go — and probably the most prevalent trend in advertising today. This kind of thinking is typically referred to as the McNamara Fallacy.

In the Vietnam War, Robert McNamara relied on a gold star metric, which was body count. As long as the number went up and the graph went to the right, McNamara had irrefutable proof that the war was being won — more people dead = less soldiers left = war closer to won. Except, as the proportion of the population claimed dead continued to grow, and the lack of war winning happened in parallel, it became clear that this single point of statistical truth was in fact false. The reasons were numerous, from officers claiming higher counts than reality to gain promotions, through to misappropriation, wrong sides, and yes, straight up lies.

McNamara assumed high quality, reliable data was being collected. When in fact what had been gathered, was nothing more than the most morbid abacus of all time.

We can attribute this to Goodhart’s law, which is, that once a measure becomes a target, it’s no longer a good measure. For example, if a salesperson is measured by the number of deals they close in a given week, soon a salesforce will be closing wherever and whatever they can. The salesperson is considered as doing their job well by closing many deals. But now that the measure has become the target (closing as many deals as they can), soon they’ll stop caring about which deals they close. Those big fish, long-cycle deals will get pushed to the side, and soon you’ll have mis-selling and all sorts. But in your metrics, your salespeople are all stars — they’ve closed everything. And yet, you’re going bust.

If you’re gauging the success and goals of your campaign, military or otherwise, by a measure — and you declare your success and make decisions based on it — you’re instantly heading down a dark path with very little at the end of it.

We see this often in the advertising landscape. As advertisers, marketers, technologists and campaign managers, we often have a desire to manage the unmanageable, and track the untrackable. We have millions of data points at our disposal, from the device you viewed an advert on through to your net worth and your social circles. But focusing on one statistic, or a group of these statistics, and setting a KPI on what these numbers should be, means that we instantly lose touch with the goal of a campaign, and confine its potential success immeasurably.

Online advertising has a tendency to be preoccupied with the performance edge — the money making attributable cogs in the wheel. And it usually will succeed in bringing you more revenue, link clicks, conversions and all the rest. But focusing only on performance advertising means you’ll never know the full potential of your brand, and means you’re focusing only on the measure. Nike would likely still sell trainers through a mix of search and Facebook ads, but would they sell as many without being affiliated with LeBron James, Tiger Woods, and those three simple words — “Just Do It.”? Unlikely. But had they never affiliated with these athletes, they’d probably think their sales figures looked amazing. They would be a smaller company, but likely still successful. Search ads and the like are wonderful tools for the performance focused advertiser — capturing the traffic at the moment purchase intent is highest. But alone, it’s like trying to knock down a wall with your bare hands. You’ll get there eventually, but you’ll go through less pain and arrive much faster if you take the time to build a sledgehammer instead.

Of course, not all performance-focused advertising mediums pretend to know all the data points — and that’s ok. Contrasted with a less techno-driven advertising medium, such as radio, it’s really night and day in terms of what is offered. Radio stations will likely have data on their listeners and the demographics that they consist of, and when the prime spots are, but beyond that, they’ve not much else to go on. They also can’t factor for talk radio Susan car sharing with R&B Sophie and her two kids, mixing things up. But unlike digital ads, they don’t pretend to know. The offering is straight-forward, you can have the radio on to make sure you hear the ad yourself, and you’ll hopefully see the impact at the other end. It’s simple, it’s easy, and comes with none of the difficulty of viewability, completed views, brand-safety and all the rest.

And that’s ok, because when a digital equivalent appears, like the Spotify ad marketplace, it’s just not the same. You’re told you can target by a host of different demographics, but are you and your buddies from work going to be humming the jingle to each other? No, because chances are, you won’t have both heard it. But you will have heard the same ad at 8am on the M25. The creative, combined with a share experience, is king.

Similarly with TV. Sure, there’s nothing to stop viewers switching over at the ads. But in the planning almost all can be controlled, and you know the advert is there to be seen — you know when it’s going live, and you can track the success at the other end with a fair degree of confidence. And you know people will watch it together — and if it’s good (or bad), they’ll talk about it. 

Yet, with digital ads, not knowing every piece of information is somehow wrong. We would rather have a number (almost any number) that is half-baked, an estimate on truth, than the transparency of saying “we don’t know for sure”. And because we take statistics to not be estimates, but the word of God, we lose sight of what we’re trying to do. If it’s a performance campaign, we lose sight of the untapped potential beyond who has converted. If it’s a brand campaign, we have no firm idea of who has viewed an ad, and what the outcome of that actually is.

Not to mention that the experience of a digital ad is incredibly insular and overly tailored.

The current monologue of an online advertiser to the audience, goes something like this:

“This is for you, you’re going to watch this before you get the video you really want, and we’ve decided this video is for you because we say so. None of your friends will like this ad, but you might. You can’t show them it, because it’s only on your device and you’ve got your headphones in. But that’s ok. Forget your social circles, the approval they give you, the shared hive mind that powers them, the in-jokes and all the rest — those don’t matter. This is our advert, you’ll like us based on your alleged interests, and that’s all that matters — you. Yes, you said you were vegan to impress your Tinder date once, but we know you, you dirty steak lover. This is your advert, and we can help you be more you. We know you. We “get” you like no-one else does. And you should be more like you and act entirely on your own preferences, because everyone wants to be their own unique snowflake at all times, and never, ever wants to fit in with the crowd. Nope. That never happens. Nope. Buy me now.”

Surprisingly, given the inherently creepy nature of targeted online ads, there’s also a distrust of the internet. Imagine that?

By contrast, TV ads still connect with audiences in a way they can trust. Between different generations, Nielsen research suggests that compared to online video, TV ads are trusted 13–21% more. And for advertisers, they know it’s a 100% whitelist. And the investment in content production is climbing up — ￡7bn in the UK alone last year. That means more shows that brands can associate with, and find the right viewing audience. Not just this, but TV advertising continues to evolve despite the mobile-first world. Taking ITVs hit Love Island as an example, which averaged 2.8 million views within 7 days of a show — and 3.5 million, once non-TV devices were factored in — a whopping 25% of viewers don’t do it live, or through the typical TV set. Online advertising, and in particular online video, need to take a leaf out of TV’s book. At present, 86% of TV campaigns lead to a positive return in the 3 years after the ad finishes. That’s why it remains a significant portion of the budget, and why time and again, advertisers come to trust it. It’s high quality, high reliability advertising.

The same, at present, cannot be said for online ads. There is a huge amount of work to be done, in terms of how viewers interact with a video, the volume of ads, the content that works, the duration an ad should be, and more. For example, at Good-Loop, we believe that the 15 seconds someone watches for before unlocking a donation is as engaging an experience as watching a TV ad for 15 seconds. It’s interactive, it’s full exposure for the brand, and it doesn’t “feel” cheap. 15–30 seconds affords brands a significant amount of time to tell a story, as it does on TV. Yet online, the trend is towards shorter, 2–3 second content snippets, or Google’s new 6 second unskippable bumper ads.

But, is it a good thing? We say that our attention spans are ever decreasing, but we can still binge watch a season of Game of Thrones. It’s the content that holds our attention, the experience. People’s attentions aren’t shorter, we just have higher production values — we demand more, and better, than ever before. And the vast majority of online video content? It’s crap.

Everything might be happening online, but if we don’t trust it, and we don’t like it, then what’s it really worth? Nothing.

Online video can be a vehicle for brand advertising, but only when it’s presented in a novel, new, and respectful way (*cough* like a Good-Loop ad *cough*). Online, is not TV. Online, is not a newspaper. Online, is not radio. It’s online. Instead of trying to provide experiences *like* TV, we must be transparent about what we’re providing. We need to communicate to the viewer in a way that is non-aggressive, non-intrusive, and non-invasive. And we need to up our game in terms of the content produced — repackaging the TV ad in a different size will no longer do.

The next huge opportunity in digital advertising, especially video, is for the effort-makers. Those that will go that tiny bit further than anyone else. The technology is in place, but in the targeting gold rush we’ve forgotten how to tell a story. We’ve forgotten how to use art in parallel with logic. It’s time for online advertising to stop the tribute act of imitating other mediums, and focus on being it’s own thing — the best own thing it can be.

This content was originally posted to Medium - click here to give it a clap :)

For more advertising ramblings, and other delights, you can follow me on Twitter here.

If you happen to have already came across this post before this share, you have my apologies :)