By Carl Carter, Marketing Strategy & Effectiveness Director, IRI

How often have you had to interact with an ‘accept all cookies’ website pop up lately? The debate around data privacy is by no means a new one, but the introduction of GDPR in 2018, and the subsequent rollout of cookie alert notices, has only increased public awareness of data privacy and how our data is being used by brands and advertisers. The rise in public concern has led to growing pressure on publishers and ad tech companies to consider data privacy in their advertising strategy and rethink their targeting methods.

This is causing many advertisers to shift away from third-party cookies - a positive step towards increasing trust with customers. And the way we use cookies could change for good following Google's announcement that it will phase out the third-party cookie on Chrome browsers by 2022.

This change has left many brands asking, ‘what’s next?’. If brands are unable to rely on third-party cookies, how can advertisers and media owners ensure they are serving relevant content to the right people at the right time?

Most are banking on first-party ID resolution or authenticated data as the answer – but this isn’t as clear cut as it may seem.

So, what is the solution?

In this post, I introduce a new approach to consumer data which I recently shared at the Festival of Marketing. The updated method of consumer targeting can increase return on ad spend (ROAS) by up to 20% by striking the balance between effectiveness (driving high volume of sales) and efficiency (minimising wastage). Leveraging permissioned first-party ‘seed’ data in a more sustainable way, the approach combines it with contextual data, sales data and predictive modelling to identify sales opportunities, and achieve reach with the optimal mix of channels, while upholding the need to protect personal information.

What are third party cookies, and what will be losing with their removal?

So, what are third party cookies, what do they offer and why have they been so important to advertisers?

Third-party cookies have a variety of uses. They are used for cross-site tracking, retargeting and ad-serving and allow advertisers to manage both interest and demographic targeting - identifying the right target audience with the right advertising. They enable a number of other functions too, including frequency capping - capturing how many times a consumer has already seen an ad so as not to annoy the user or waste investment on over-exposure. These cookies also allow for advert retargeting – which commonly occurs when you view a website, or a particular product online, such as shoes or clothing, and then see display ads for those items across other websites you visit.

Possibly most importantly, third-party cookies enable the process of connecting multiple identifiers across different devices and platforms, and the individual data attributes attached to those identifiers, to create a single, omnichannel view of a customer. This allows advertisers to map consumers across different sites and track the many touch points in a shopper’s journey. For instance, perhaps your customer viewed your product on your website from a display ad, but they leave your site and later purchase the same product on Amazon.

So, we can see that the decline of cookies poses an issue for advertisers. But in a data-aware age, privacy and consumer trust are more important for brands. Change is a good thing, but brands need to find a new way.

So what is first-party data, and can it solve those challenges?

Let’s start with the concept of identity resolution. It's the idea of taking first-party data, like an email address or household address and matching that person across multiple partners or publishers. Essentially, it’s creating a unique ID that allows a brand to know who a consumer is, in an anonymous privacy safe way.

It opens up the doors to a direct relationship with the consumer who has given permission for you to have their data, which is immediately a more trusting and open relationship between the consumer and the brand. This means you can still target customers with similar approaches discussed above, but arguably in a much stronger way because you develop a one-to-one personalised relationship. The use of this data to activate connected media also means we can measure the impact of advertising through closed loop measurement techniques, such as comparing those who are exposed to the non-exposed, those that haven’t seen your campaign.

The possible downside, however, is scale. Even in strong cases you might only be able to deliver a match rate of 20% across your first-party data set –so with one million users you may only end up matching 200,000 users. If you’re an FMCG brand, activating against an audience of this size isn’t likely to drive strong incrementality or ROAS . Therefore, it’s really important for brands to have first-party data at scale to provide the strongest seed of deterministic data.

What does this all mean for brands who need to drive volumes sales?

This leads us to the final piece in the jigsaw, predictive modelling. IRI analyses the world’s largest purchase data set direct from source to help FMCG and media companies grow their businesses. We use the most impactful KPIs from retail, such as sales value, units and product distribution at the most granular level providing vital signal data.

We overlay that first-party data with contextual data sources to help brands develop purchase-based planning and targeting, to determine a granular view of high and low opportunity areas. For example, this can highlight areas where sales are saturated at the brand level or the category itself is in decline. It also provides a view of high opportunity areas that can drive significant penetration and sales for a brand.

At a geographic level IRI integrates with data partners such as Adsquare, Eyeota, LiveRamp or direct from demand side platforms (DSPs) and blends this with a brand’s first-party audience, as well as second-party data to model look-a-likes for increased reach. This process creates a unified dataset to allow brands to significantly grow their audience and thus chances to drive incremental sales at a volume that is impactful.

Unifying the data layers further provides greater accuracy, relevance and scale

Starting with purchase data provides the most robust signal to build out a holistic targeted advertising model.

For example, household and psychographic data can identify types of consumers, household composition, values and ethics. Location data can indicate shops they may visit and locations they spend their time such as gyms, coffee shops or DIY stores. By modelling this data together at a geographical or audience level, we can increase the accuracy and scale from first-party sources.

This allows advertisers to understand the key attributes in their customer base and their customer’s journey, all without infringing on consumer privacy. The approach allows advertisers to balance the vital mix of efficiency and effectiveness because they can qualify high potential on-target audiences to be efficient whilst delivering the high volume impact for effective media contribution.

There are many challenges facing advertisers in a post-cookie world. Cookies have been such a valuable tool for marketers, but there are many advantages to first-party data, such as greater quality and detail. First-party data alone however does have its limitations, such as the difficulties of achieving scale and reach, and the challenge for brands to connect all media partners’ data sets.

IRI’s layered approach enables a more sustainable way to leverage permissioned first party data and reach business goals. Using data modelling and combining the optimal mix of channels will deliver efficiency and effectiveness for advertisers beyond the cookie.