Blog Online Identity in a Programmatic Age

Online Identity in a Programmatic Age

Posted on Mar 12, 2019

A Platform161 Pocket Guide to how user identity across digital platforms works right now, and how it might in future.

Historically, advertising was all about speaking one-to-many. But since then, we’ve witnessed the huge shift from a handful of print and analogue TV options offline, to a multitude for consumers online. And with this explosion in choice, and fragmentation of how we spend our free time, the era of one-to-one marketing at scale was born.

Originally, each magazine or even TV show stood as a fixed representative of a particular type of person or audience group. But since digital has taken over, we’ve moved closer and closer to marketing based on an actual, live view of what our audiences really enjoy or want to buy in the moment.

From paper coupon to plastic loyalty card to virtual cookie, smart brands have always known the importance of understanding the customer. It’s just that the options around developing that insight, along with ad buying choices in general, have expanded immeasurably.

In a world where we consume media across multiple screens, and multiple locations, the potential benefits of getting identity right are greater than ever. Many of those benefits are also tied to the original promise of digital. That is, making advertising more measurable, enabling accurate frequency capping, and even sequential and creative messaging – and that in turn all leading, at least in theory, to more impact, better results and lower wastage. But of course, as with everything in the online economy, there isn’t only one way to approach digital ID. We’ll start by reassessing the original approach, which in one form or another has been around for the past 20 years.


The cookie has proved a remarkably durable way of building and measuring campaigns. Some say it becomes less effective in newer environments – for instance, measuring activity within mobile apps can be a challenge. And equally, cookies weren’t necessarily built to understand that Jane on her laptop is also Jane when she is on her smartphone, or even watching her smart TV. Mobile will make up more than 70% of the US digital ad market in 2019 (eMarketer), so you can see the imperative for getting cross-screen measurement correct.

Without a doubt, cookies are and will continue to be a massively useful tool to advertisers worldwide – but increasingly, alongside other tools and technologies – when it comes to campaign tracking in particular, postback tracking is one alternative to cookies that’s being presented. In other quarters, it’s those more focused around what’s known as ‘people-based marketing’ where people are focusing efforts.

“what we’re ultimately looking for is something much like the old loyalty card or loyalty scheme – but in a digital, multiscreen context.”


When we speak about cookies, as above, we’re often talking about 3rd party data sources, and ‘syncing’ data to get a better understanding of the user. Platforms then deduce information about them – ‘probabilistically’ attributing likely interests or demographics.

Where people-based marketing generally differs, is that instead of 3rd party data, its starting point is 1st party – more likely trustworthy, user-disclosed signals. From this – so-called ‘deterministic’ rather than probable source – the idea is we have an even more accurate picture of our audience, across more (if not all) of the devices they use. People-based marketing is based on robust insight on a person. And campaigns are built on that framework, instead of the starting point being an estimate, or likely sign of intent.

The potential benefits don’t end there though – add to the list lower chances of fraud, higher viewability, more accurate cross-device measurement, as well as improved frequency capping.


Another potential, and interesting solution for the future of people-based marketing may lie in the various ID consortia that are currently being brought to market. One of these is the IAB-owned Digitrust. In theory, they promise to level the playing field, with various ad tech companies sharing the same identifier system, while also ensuring users’ data are only shared as they want it to be. The ultimate aim is to collectively hold a better, more accurate picture of the consumer. And also, in this way, present a more viable alternative to Google and Facebook – who can both already do this with ease. And also, to cut down on user syncing, and thereby also latency, improving the web experience for users too.

So, the consortia’s potential is great. On the other side, they are also largely still in construction, or trying to reach critical mass. Some might commit you to using certain ad tech partners over others. And by extension, become more reliant on their own partners. Additionally, more persistent cookie-like ID solutions could be blocked by browsers in future.

Expanding the list of 3rd parties you’re reliant on can increase overall cost and even lead to further 3rd party requirements. For example, some brands already need to work with a partner dedicated to ‘data onboarding’ – that is, just to connect all of their own data to various marketing applications.

In some ways, this takes us back to the original vision and benefits of people-based marketing – and a single data source independent of multiple external parties, some unknown to you directly.

When building campaigns and buying media, accuracy is everything. Especially for identity, location or any other key factor. Think of it like the foundations of the house you’re building. One piece of research from Mediasmith tested eleven data providers, and found four of them delivered performance that wasn’t much better than chance – in other words guessing users’ age and gender at random.

“People-based marketing is based on robust insight on a person. And campaigns are built on that framework, instead of the starting point being an estimate, or likely sign of intent.”

What this points to, as we’ve referenced before, is the fact that we have different types of data – probabilistic and deterministic. And deterministic is preferable to most other types, not just because it should be more accurate. Its other advantage versus 3rd party data is around legislation like Europe’s GDPR – if similar measures get adopted in other countries, 3rd party data sources built on probabilistic methods may become less sustainable at some point in future.

As we move from 1st to 3rd party data, we’re on an inverse sliding scale of audience size increasing versus accuracy potentially going down – which means of course that further measures must be taken to monitor third party sources’ truthfulness.

When it comes to that precision, what we’re ultimately looking for is something much like the old loyalty card or loyalty scheme – but in a digital, multiscreen context.

But no one has fully succeeded in creating anything matching those solutions in a digital context. At least, not as yet. Ultimately, that is still both the challenge, as well as the great promise of all the various ID schemes currently vying for supremacy in programmatic advertising.