A Guide to Buying 3rd Party Audience Data

While everyone knows that programmatic display sits at the top of the digital marketing funnel, what some people are less aware of is that there’s an intent funnel going on within programmatic display which should be realised to create any solid campaigns. Yes, no two campaigns are alike, but whether your success metrics are viewability or CPA, you need to understand how to effectively buy and optimise 3rd party audience data to achieve and surpass your targets.

What is 3rd Party Audience Targeting?

Before diving in, it’s worth recapping exactly what is involved in buying 3rd party audience data segments. First of all you have the 3rd party data providers. The household names include BlueKai (acquired by Oracle in 2014), eXelate and Lotame to name just a few. In actual fact there are, at the time of writing, 49 different companies whose core business is scraping user data from online (and offline) sources and packaging them up (within their DMP) into similar behaviours or ‘segments’ for ad buyers to purchase. At this point it’s always worth adding: All these audience segments are cookie-based and anonymous… PHEW!

Now, in the vast majority of cases, these segments will be made up of a mixture between declared data (eg. a user has explicitly entered information about themselves), behavioural data (what articles a user has viewed) and inferred data (using the DMP to model or scale up the segment using a lookalike algorithm). The blend of the three data types above can have a huge impact on the quality of the audience in question. This is one of the biggest challenges faced by ad buyers in the current programmatic climate as there’s still a lack of transparency when buying through a DSP (Demand Side Platform) as to how that audience segment was built.

The 3rd Party Audience Data Market

However, 3rd Party Data Providers are going through an explosion of sorts. New companies are cropping up almost monthly who believe they can better source the user data, collate more accurate and representative segments, and apply better algorithms to scale up these audiences using lookalike modelling within their own, proprietary DMP. Therefore, there is suddenly a lot of variation in audience segments. Some data providers are focussing on different verticals (eg. Eyeota’s partnership with Bombora for B2B advertisers) or focussing on different parts of the intent funnel (eg. Google’s ‘In-Market’ Segments) I mentioned earlier.

Now it’s worth briefly describing this funnel. All programmatic display campaigns (and probably all advertising campaigns) fall on a scale which at one end is pure brand awareness and at the other is pure direct response:

For campaigns focussing on the former, Account Managers are looking to cast their programmatic nets wide but also target the right sort of user – this would be where demographic targeting, for instance, would be most applicable. The aim of the game is to expose your brand to users who probably don’t need the service yet! But because you’re targeting the right kind of users based on such attributes like income, age, gender, marital status etc., when they do need your company’s services, it’s your brand they will think of first.

Direct response is all about finding the users who are already in-market, window shopping for the service or product your company offers. Buying these audiences means you’re forgoing reach for accuracy and you’d be expected to need more, smaller sized segments in your targeting arsenal. Ideally you’d forgo any lookalike modelling from 3rd party data providers and have an audience list full of users who have carried out the Google searches, article views and visits to competitor websites which signals they’re “ready to buy”. The diagram below demonstrates how you should want to buy data based on your campaign goals:

Measuring Success

So combining the “Programmatic Display Scale” with the above funnel you can begin to understand how best to target your audience based on the primary campaign goal (and how best to measure success!). It’s therefore paramount that these core goals are set and adhered to throughout the campaign. Time and time again we see advertisers who muddle their primary goals and want one campaign to do it all so it’s no wonder they get disillusioned when CPAs rise due to increased brand awareness (Top of Funnel) activity and deem the whole campaign to be a failure!

Hopefully this article has solidified two important areas. The first one is how a campaign manager should buy and target audiences. But in order to achieve success with third party data in the first place, clear campaign goals (which cannot be “all of them”) must be set in stone from the off.

If you have any further questions about how Periscopix approach programmatic display campaigns, on 3rd party audience data or anything else please view this blog written by James Dix on Periscopix's website and don't hesitate to get in touch.