More data does not always mean more effective marketing

Media buying is trending towards hyper-personalisation, but sometimes broader targeting through high-quality media will reach more potential customers.

Data segmentation
Source: Shutterstock

“You know, I think we’ve already got enough consumer data, introduced enough personalisation, and leveraged enough advanced adtech in our programmatic media buying.” That’s certainly not something you hear very often in marketing circles.

In fact it remains broadly taboo within the industry to suggest that we’re on anything but an inevitable march towards hyper-personalised and -targeted communications, in which owning consumers’ data is as valuable as their actual custom. To suggest anything else marks you out as a backwards Luddite who does not understand the power of new technologies.

I’ve seen many a senior global marketing leader proudly espouse the importance of such levers within their own businesses strategy, and sign off millions of dollars accordingly, though it’s hard to know to what extent they truly believe this or are just going along with it so they don’t show themselves up to be behind the times.

One of the biggest motivations I’ve seen is a universal sense that the competition is playing heavily in this space and your brand needs to catch up. Five years ago I wrote an article which compared marketing team conversations about data with high-school conversations about sex – everyone’s talking about it, but hardly anybody is doing much of it well.

It remains broadly taboo within the industry to suggest that we’re on anything but an inevitable march towards hyper-personalised and -targeted communications.

A few times whilst I worked at GSK I remember a shockwave rumbling through the business, with board members, the C-suite and everyone around them suddenly asking extra questions about our data strategy. Typically, it was triggered by a comment in the press or on an earnings call, by a company like P&G that was making bold claims about its own data approach. Before you knew it, absolutely everything we did needed to focus on data acquisition, targeting and personalisation.

It’s certainly true that P&G has an impressive approach to data. On some of its brands, such as Pampers, it clearly has incredibly detailed data strategies and customer journeys that look to follow families from the moment of birth through every step of their kids’ development. When it comes to selling Tide laundry detergent to most of the US population, however, you can imagine the brand is less concerned about data sign-ups and hyper-targeting.

Less can be more

I’m here to be the therapist who tells you that, whilst clearly sometimes these things are important, it’s also absolutely fine to sometimes say you don’t need so much of them. Even personally, in my own roles, I wished I’d had more confidence to speak out against the tide when I didn’t really think some strategic decisions really matched what would unlock the most value for the business.

At the end of 2023, a Programmatic Supply Chain Transparency report by the US-based ANA showed that about one third of the money advertisers spend on programmatic media makes it through to the final consumer and publisher. A slightly higher proportion is completely lost in the system to impressions which aren’t viewable, are fraudulent or end up on low-quality clickbait sites.

The final third goes knowingly to the adtech and data which enable your programmatic buying, and may indeed add enough value to justify its existence in the mix. By the time you’ve added in agency or tagging fees to the waterfall, it can easily be less than 20p in every pound actually ending up buying you space with a publisher.

Five trends that will shape media in 2024

In most supply chains, such leakage would be a shocking revelation, yet in media it only caused the slightest of ripples. Perhaps that’s because the in-depth report went into more than a dozen nuanced implications that were simply too much to take in. Perhaps it was because the numbers weren’t really anything new and simply echoed similar results from ISBA a couple of years previously. Perhaps it echoes a broader blind belief that these technologies must ultimately still be worth it because everyone else is using them. You’d think, one way or another, that the $22bn in potential efficiency gains they identified might have grabbed some attention.

If you take the two thirds of your media budget that would otherwise be used up in the system and give that directly to a media owner, they’ll be more than happy to offer you substantially better value, reach and likely some bonuses on top. A deep dive into supply path optimisation of your own media approach can help identify the partners and technologies worth their weight in gold and inclusion in your plans, and where you’d do well to patch up your leaky bucket.

The sunk costs of data

In the same vein, the idea of acquiring as much first-party data as possible to better understand your consumers and to more finely target your communications is hugely alluring. The discussion of it, however, rarely factors in the true costs of collecting that data, let alone the large sunk costs in CDPs and data infrastructure to collate and store it. The kicker? The more data you add into your media buy the higher, the cost you’ll be paying for every thousand impressions to reach the right consumers.

That’s before you factor in how quickly data ages, becomes irrelevant or ceases to match up with the technology you’re trying to leverage it through. And that’s also before you start to ask questions about how accurate data ever is to begin with, or how easily misleading the signals we all give out can be.

Consistently in my work at companies including Mondelēz, Diageo and GSK, I’ve seen examples where, by the time you factor in all of these costs, it would have been cheaper to reach your hyper-targeted audience simply by carrying out a broader media buy that happened to included them, as well as countless others as a bonus.

The more data you add into your media buy the higher, the cost you’ll be paying for every thousand impressions to reach the right consumers.

It’s a hugely tempting prospect to reach exactly the right consumer, at the right time, with the right message. Most marketers aren’t going to say no to that rich promise. The challenge is we don’t often discuss enough whether reaching quite a lot more of roughly the right consumers at almost the right time might sometimes just be more effective. Especially if some of those extra eyeballs you ‘accidentally’ reach end up being in the market for your brand some months later.

One of the very first mentions you can find online of “personal connections at scale” is a Guardian article I wrote back in 2015 about some of the work we were doing with Mondelēz and Facebook. Almost 10 years on, this notion of moving towards highly individualised communication remains a North Star for many businesses, but if taken in its most literal sense, there is little evidence to suggest it’s actually a hugely effective marketing tactic.

That’s partly because it’s often done badly, partly because it’s hard to get the simple economics of targeting versus reach right, but also partly because it does something of a disservice to the power of big, unifying creative ideas.

A balance of possibilities

I happen to still personally believe in the huge potential of following an approach that delivers personal connections at scale, but I think there is plenty of rich middle ground for businesses to explore away from the extremes; an approach which is certainly more nuanced than simply running one TV ad, but doesn’t actually require individualised messages for every single consumer. Many brands haven’t truly experimented in basic personalisation, such as exploring different passion points through contextual targeted creative, or talking relevantly to large diverse populations through specifically targeted publications and news media.

None of this means that data-driven marketing, personalisation at scale or programmatic media are bad things, but it does mean that we should always be asking questions about whether they’re the right thing for the specific campaign or marketing objective we have in mind.

In particular as marketers, and especially as media experts, we need to understand that the answer is almost certainly a balance. These new technologies and data approaches are absolutely worth exploring, but the true test of something isn’t just whether it works in its own right, but whether it worked sufficiently better than other uses of the same money.

Marketers prioritise own data as ‘illusion of control’ from third-party cookies vanishes

Despite the moves by traditional media channels to also allow themselves to be bought programmatically, it’s far from inevitable that all media will one day be bought this way. Many of the more advanced companies I speak to through the WFA, or now my own consulting work, have in fact reduced the role that programmatic media plays in their mix in recent years. Perhaps even a majority of larger companies are turning their backs on truly ‘open market’ programmatic, which can see campaigns run across tens of thousands of unknown sites, and moving to inclusion lists and private marketplaces.

You’ll struggle to find articles or presentations about them, but there are far more horror stories about failed CDPs and data strategies behind closed doors than there are tales of overwhelming success – though in the right business conditions those certainly do also exist. This piece exploring the true value of first-party data is certainly worth a read for those wanting a deeper dive.

It’s fair to say that AI and other emerging technologies are going to continue disrupting this space over the coming years. Much of that will push us closer towards a world in which the benefits of advanced buying and data approaches outweigh their costs and cautions. That just isn’t always the case yet.

It may seem a little old school, it may make you worried that you’re old-fashioned and missing some trick, but it’s OK to question how much data and targeting you truly need to get the best results. It’s also OK to conclude, with the right evidence, that the answer for your business may well be that it makes sense to use an awful lot, especially if you work extensively with performance marketing. You can also understand that data and targeting might be more relevant for a highly niche product (for instance a very premium whisky) than it is for something more mainstream (like a mainstream bourbon).

The great thing about looking back towards some more ‘traditional’ approaches to media – which prioritise partnerships, known media titles and high-quality context – is that you’re also side stepping many of the wider issues that digital media faces. You’ll hardly notice the death of the cookie if you’re relying more on broader audiences or native partner targeting. Brand safety, media quality, viewability and the like can be managed by taking greater control of your campaign governance and setup, but without any of the more substantial risks that the open web entails. A lot of this falls very much in line with the new WFA Global Media Charter, and its challenge to advertisers to rethink their media approaches and ecosystem over the coming years.

Currently being a free agent does enable me to write a perspective like this without encountering the wrath of my corporate comms team, but for the sake of my future employers I do feel I should end by reinforcing that I truly believe in the power of data and programmatic when done right. So much so, in fact, that I once helped create a training workshop called ‘Programmagic’. However this avoided getting lost in the technicalities of possibility, pushing marketers to ask what was strategically or creatively valuable, which we should then push the tech to deliver for us.

Jerry Daykin is a global media leader who has worked for companies including GSK, Diageo and Mondelēz. He is also a WFA Diversity Ambassador.