Brands have previously found themselves in something of a Catch 22 when trying to maximise the usefulness of multiple data sets.
Historically, creating actionable insights out of disparate and siloed sets, both offline and online, by harmonising that data to create a more unified view, meant the data had to leave a company’s systems to be matched by a third-party, risking trust and privacy challenges. But if they didn’t onboard and match their data elsewhere, they risked losing competitiveness by not being able to maximise the information within that data. Tricky.
The Holy Grail is a solution where individual-level insights could be obtained with zero risk to data security, which technology innovation is now making increasingly possible. But the ability to do this is only the start, brands need to understand the enormous possibilities for marketing once they can create maximum value from their data with zero privacy risk, if this new model is to leapfrog the old.
Why do brands need to onboard their first-party data?
Quite simply, brands need to onboard their data to enable a 360 view of what they really have. Most brands have manifold pots of data, some of which are offline, such as CRM entries, and likely to contain personal data identifiers such as names and addresses, phone numbers or email addresses. They are increasingly likely to also have a wealth of online or digital data, such as browsing and conversion data, for example, information generated through using cookies or mobile IDs for tracking. This is generally pseudonymous, identifying at cookie level rather than through direct identifiers (such as name or postal address). But brands lack the capability to easily bring these two sets of data together.
Data onboarding, usually via a third-party ‘Identity’ vendor, allows brands to combine these data silos, giving them a more holistic view of the data - and, therefore, of their customers. It also allows them to augment this new knowledge with other external datasets. For example, onboarding customer data to the walled gardens of advertising platforms such as Google allows brands to combine information from a search term with what they already know about that customer. Take car insurance firms - being able to identify someone searching for ‘car insurance’ and knowing whether they are an existing or prospective customer is incredibly useful in determining what the best offering for them is likely to be.
It also allows more cost-effective bidding within search results, because the companies will have a better idea of when to bid low, if they are an existing customer whose insurance is not up for renewal imminently, for example, or higher if it is closing in. In short, it allows better contextualising and a more relevant offer. Another benefit of data onboarding is the reduction of irrelevant retargeting, because a retail outlet, for example, would be able to tell from its combined data that you later bought the shoes you looked at online in one of its stores, reducing the numbers of adverts that follow all of us around the web.
Onboarding in order to combine data sets, then, also offers brands more control over the content and frequency of their marketing messages, by having a more rounded view of their end customer.
So far, so good - but there have been problems with this approach. Most notably that brands have to physically share their data outside their own environment, with all the potential risks this brings. The use of third-party vendors, too, who are not as recognisable as the brands using them, can also cause alarm bells to ring with increasingly privacy-conscious consumers.
Additionally, because most Identity vendors are not walled garden behemoths like Google and Facebook, which have access to immense amounts of user data, the scale of their identity graphs - and thus the depth of insight they can offer to brands from the data that has been shared - is significantly curtailed. They may not also operate in every country, which again means fractured marketing, where the right hand of offline data quite literally does know what the left hand of digital data is doing, continues. Far from ideal.
Google and Facebook, of course, both have their own hugely successful offerings in this area - Customer Match and Custom Audiences, respectively - but these also require data to be physically shared.
However, there is another way - an alternative to achieve the maximum possible insights without using these walled gardens, which is gaining ground at pace. Let’s call this the 2020 approach for data onboarding - and you won’t be surprised to learn that InfoSum is one of the pioneers.
2020 privacy-first data onboarding
The InfoSum platform - and we truly believe this is the future of onboarding - allows brands and media owners to gain huge amounts of insights from their data without physically sharing it with a third-party. Instead, mathematical representations of the content of individual data bunkers, which remain entirely under the data owner’s control, are used to understand the intersection - or not - between a brand and a publisher’s combined offline and online data sets, for example, without either having to physically hand over that data.
Because of the cutting-edge analysis tech which is used to analyse these data representations, we can also extract more from it, comparing not only deterministic data such as addresses or email addresses, but also looking at the data probabilistically, and assessing the chances of different pieces of data, such as name and postcode, held separately within the data, actually being the same person. This results in a superior match rate, which in turn leads to more precise marketing - all without any personal data being actively shared. This non-movement of data also reduces the compliance burden, which is itself important as privacy regulation increases globally, most recently with the CCPA in the US.
The future of first-party data onboarding
So what does the future of data onboarding look like? It looks like a true win-win, providing all of the benefits of merging data sets, notably actionable in-depth insights, and the ability to know your customer better and more deeply, without any of the security or privacy risks of actively sharing data. Best practice in data onboarding in 2020 means activating your data with maximum impact, being able to sift your customers more accurately and consequently spending less money advertising with better results. Who wouldn’t want a slice of that?