Data, data, every where, Nor any drop to drink. #DigitalSense

Here lie 8 key questions anyone in marketing should be asking about their data strategy & approach to ensure it truly makes digital sense. 

These days as a marketer (or indeed as a consumer) we are permanently surrounded by data. By people talking about data, by people selling us data, by people using our data, and increasingly people worrying about their data. Yet there’s a very real danger amongst this vast data lake that, like the poor Ancient Mariner stranded on the salty sea, we can find that the more data we drink the more we need and yet our thirst is still not at all quenched.

An unpopular truth about data is that over the last decade it has made marketing demonstrably worse. Sure, some companies only exist because of data, and others have found the balance in using it to truly step up the performance of their business. The work of Binet & Field (and indeed others) has however shown that overall, during this new era of data-driven digital marketing, effectiveness has gone down - and most assuredly so for big brand building advertisers mislead by a short term opportunity.

Not the admission you normally hear at the start of an article on data is it? Yet refusing to acknowledge the risks of data is a guarantee that you too are heading down the wrong track. Data can be expensive, data can lead to over targeting & the loss of growth audiences, data can actively exclude minorities, data can defund journalism & spread fake news, and crucially data can cause us to lose focus on the importance of great creativity as well as the value of broader fame & cultural significance. It may surprise you to hear that most substantial research (eg Ehrenberg Bass) into the actual marketing effectiveness of hyper personalisation (and I don’t mean just the media efficiency thereof) has shown little or no benefit, or certainly less benefit than from having a universally well received creative.

Of course, when well-tamed and well-leveraged data can also be of huge benefit to marketers, and you’d be a fool not to be exploring the possible upsides for your business. If you can use it to be relevant to more people, not just to be visible to fewer, then you’re onto a good thing. A well implemented data strategy uses that very data to know when your marketing needs to go big & wide, but also when to narrow in on a key target.

The future of data itself is a murky one - legislation, technical platform changes, as well as shifting consumer perceptions will all limit the amount of data available to marketers. It’s highly probable that the future of marketing is a future with less data available than we have today, but that makes it even more important to use this time to work out when it pays off, and when it doesn’t.

So as we embark on this data journey together here are some of the key questions we all should be asking - and indeed within my own team I’m pushing this as the EMEA Year of Data, a key time to build capability before the cookieless future arrives.

Is it targeting? There’s a broken assumption that effectively leveraging data takes us straight to targeting, or even hyper targeting and personalisation. Not at all - as has been the case for decades some of the greatest value of data, now in new digital forms, is the overall strategic guidance it gives us as we shape our approaches and our creative. Identifying that gamers are a potential growth audience for your brand could be a great insight, but it doesn’t mean you need to launch into some hyper targeted response - perhaps the answer is in a broad creative message that cuts through to gamers wherever they happen to see it, perhaps it’s bigger partnerships or product development you need. Some of the best uses of data for big brands are to help plan and inform critical mass reach tools like our TV plans and OOH strategy, far from the digital front line we often jump to.

Is it needed? There’s been a rapidly spreading notion that it’s ‘best practice’ to apply lots of data to our media buys, especially in programmatic spaces and even on social. Allegedly you’re not doing good marketing or making the most of a platform if you aren’t piling layers of data on. Data can indeed help you more efficiently reach a very specific target, but ultimately we know brands grow by building penetration across much lighter buyers of their product - the kind of audience who might very well give off no identifying data signals at all. On big platforms like Facebook in particular there’s a lot of evidence that very broad targeting ends up being far more effective overall, not least because it continues to reach your most core audience whilst also expanding beyond it. Greater use of data & targeting can make more sense when you’re looking to move people further along a consumer journey though, or have a very specific of niche product to sell so it’s a question of evaluating the strategic requirements and true need - and asking whether other tricks (such as basic context which we often seem to forget) could do the same. Certainly some basic constraints aligned with your potential target market, eg geo location for a local restaurant business, can go a long way to reducing wastage.

Is it accurate? That’s a question that is surprisingly rarely asked, and which with brutal honesty few of the key players in the ecosystem have much incentive to tackle. The data advocates in your own marketing team, the agency helping source the data, and certainly the company selling it are all pretty happy not to ask. We tend to have an underlying assumption that the data we are using is very accurate, after all if we’re trying to use it for very specific activities it would need to be. The reality can often be quite different - just look at the ads targeting data platforms like Facebook & Twitter have on you. Certainly some of it may sound spot on, but there’s usually about as much which doesn’t sound like you at all... what did you click to make them think THAT? There are often ways of carrying out validation if you really try - partners like Nielsen who can help verify on target reach, or like Location Sciences who explore the accuracy of location data. You won’t always like what you find, but you can use it to pick the right partners or negotiate for fairer value.

Is it worth it? Even if all the stars align there’s still a fundamental question of value and rarely enough rigour in this space. Data’s validation usually comes through relatively self-fulfilling performance metrics, and certainly it might help you buy a specific audience more efficiently, but it takes bigger analysis to work out if it was effective. Crucially there’s no such thing as free data - sometimes you literally pay an upfront charge for data, but even if the data is provided free from a partner (or you are able to leverage your own) then there are hidden costs. You might for instance face a higher CPM triggered by a narrower search, or the opportunity cost of the audience you’ve now excluded. When best implemented, data can help you beat your competition in an auction by understanding that a certain impression is worth more than they realise, and certainly with a limited budget data can often make your media work harder. The real test of data is not whether it’s performed well in itself however, but whether it outperforms the results you would have had if you put all the same effort, budget and attention into improving the creative itself & showing it to a broader audience.

Is it self-fulfilling? The use cases for data, and seemingly the value there-of, increase as I’ve said as you move down the purchase funnel and especially as you look at highly engaged category buyers. The real question here is whether you are cleverly intercepting or protecting a valuable consumer that a competitor would otherwise have picked up, or whether you’re just paying a tax to convert people who were headed your way anyway. Done well this is a critical and valuable part of modern marketing, but even deep into the funnel all the evidence shows that the biggest advantage and boost to your data approach is having a powerful and well established long term brand behind it. Sometimes the focus on this conversion can blind us to another data opportunity too - “How do I use data to reach an audience that really isn’t interested in me?” is often a much richer brief than “how to use it to reach those that already are?”.

Is it here to stay? I’ve hinted at the many factors which will limit access to data in the future, the dawning of a ‘cookieless world’ and a new dawn in personal privacy awareness & legislation chief amongst them. Before you invest too heavily in a data strategy you need to ask some tough questions about whether the approach you have in mind will even be possible in a year or two’s time. In Europe we’re still to feel the real impacts of GDPR legislation, but we already see far greater caution and control around data usage. The ecosystem is adapting around these challenges and throwing up new solutions so it’s far from the end, though many of which rightly or wrongly play into the hands of big data rich companies like Google, Facebook & Amazon. We should see a resurgence in the lost art of context too, sometimes the easiest way to reach a football fan is to advertise on a football news story, not to use data about their recent purchases to chase them around the internet.

Is it right? Not just whether you have the legal rights to do something, but whether that something is right to do. Data ethics is a nascent field in marketing but a critical one which pushes us not to just ask what can we as marketers get out of using data, but what is the fair value & exchange with consumers, and how do we use as little as possible of their data to bring as much value. Public trust and opinion of advertising has declined radically over the past decade, largely driven by misuses of data leading to adverts chasing us around the internet or seemingly knowing too much about us. If marketers want to be able to use some degree of data going forward (and the basics are needed to cap frequency, stop fraud and to have any idea of what’s happening) then we need to commit rapidly to making better and fairer use of the data we do have. With the news filled with data breaches and other horror stories there’s also a question about how the industry better makes the case for the positive value of data.

Is it inclusive? Finally, it’s always worth flipping your data strategy on its head and asking not only who is it focussing you in on or targeting, but also who as a result are you excluding. There are many cases in which blunt application of data can exclude a range of minorities from your advertising in particular and whilst missing out on a commercial is hardly a hate crime, it can manifest itself into missing out on opportunities or even the defunding & silencing of those minority voices which might be. Crucially from a business perspective it can also often mean excluding potential growth audiences - need states and category entry points you perhaps don’t yet fully understand but who are included when you go broad with your advertising, but whom you become invisible to if you go too narrow.

Despite the questions and doubts above I remain a strong digital evangelist and an exponent for the power of data in driving the future of marketing, but as an industry we need to ask as many hard questions about why & when we shouldn’t be using data as we ask about how & why we should - only then will we find the right balance for our brands and for our society.