More and more companies are trying to get into the people-based marketing space. But while many claim to have data sources, they simply rely on device IDs, not building accurate and in depth personas for individuals. Toby Benjamin, VP Platform Partnerships, Viant, explains how they’re missing the three key things which constitute a true people-based platform: Scale, accuracy and depth.
People-based marketing is the hot term in the advertising industry right now.
Lots of companies are talking about the benefits it can reap for brands and the impact it can make to their bottom line.
This is great – more conversation around making advertising more accurate, personalised and relevant has to be a good thing.
But, in my view, some companies claiming to offer people-based marketing are not what they seem. And with so many jumping on the bandwagon, there’s a real risk that advertisers are over-promised and under-delivered results for their campaigns. Which could ultimately undermine their confidence in a new advertising strategy that is successfully moving the needle for many brands.
So how do you separate the fact from the fiction when it comes to people-based marketing? For me, there are three factors that underpin a successful people-based approach: accuracy, density and scale. Asking the right questions about these will help you weed out the fakers.
1. Do you really have accurate data?
The power of a people-based approach comes from knowing an individual’s identity (e.g. ‘I know this is John Smith’) and being able to target with precision (e.g. ‘I know John Smith is also a Renault driver’).
A people-based provider will deliver this accuracy based on first-party data. By owning that data and having opted-in registered users, a provider can then match this data with companies offline to enrich what they currently know about those customers. For example, a partnership with the RAC will give brands access to exactly which vehicle a person owns, rather than relying on guesswork based on online behaviours and sites they’ve visited in the last few days.
This is very different from the traditional cookie-based approach, which relies on making assumptions and inferences based on a user’s behaviour (e.g. ‘this user visits Renault blogs, so could be a Renault driver’).
A few questions around accuracy should help weed out the fakers:
• Does the provider have access to a first-party database, and the consent to market to those individuals directly?
• How often is your data refreshed? Obviously people’s circumstances change and so do their likes and dislikes. The fresher the data, the more accurate it will be.
• Do you have real deterministic audience insight (good) or is it modelled or probabilistic information (not so good)?
2. Do you have data density?
People are complicated. We are all individuals with a myriad of likes, dislikes, and behaviours that are unique to us. That means for advertising to be truly relevant to an individual, you need to understand every part of their lives. When you have true first party data, you can generate partnerships with other offline data companies that can give you insights on individuals like what car they drive, where they shop, where they like to go on holiday, what they like on social media etc.
This data is most powerful when you bring it all together with a people-based approach – giving what I call data density.
Data density lets you build really meaningful audience segments to target, rather than relying on broad brush stereotypes. A people-based faker may claim to be able to do this, but their data won’t be of a high quality, as often it is based on a set of assumptions.
So try asking:
• Where are you getting your additional data insight from? If the partners are not high quality sources you trust, alarm balls should be ringing.
• How are you matching different data sets when you bring them together – is it identity-based and a one to one match (good), or are you matching broader categories like postcodes or modelling out from panel data (not so good)?
Do you have true scale?
The third factor to consider is scale. There is no point having accuracy and density if you end up with an audience segment too small to target in a way that will make a meaningful difference to your business.
In fact, there are only a handful of companies that can really achieve people-based scale. Companies that don’t have millions, or even billions, of first party registered users will be relying on modelling an audience with cookies or device IDs to achieve scale. This will degrade the quality of your data, as you’re no longer relying solely on first-party information to identify real individuals but finding lookalike audiences based on probabilistic behaviours.
So if your provider is saying they can have this kind of scale, but they’re not a company you’ve heard of and can’t explain how much first-party data they have then that should raise a red flag.
Try asking:
• How are you accessing that many individuals?
• Do you mash together cookie and first party data? If they are accuracy is lost.
• Are you using modelling and assumptions to build scale? If they are, you will lose quality.
In future, expect even more companies to be talking about people-based marketing. And if you want to avoid the fakers and make sure your people-based marketing is a success, be sure to ask about the three areas above.
By Toby Benjamin
VP Platform Partnerships
Viant
http://www.viantinc.co.uk/