How AI-driven predictive analytics will benefit marketers in the new data economy
Marketers are all too aware of the issues that are driving us towards a new data economy. Google’s restrictions on the third-party cookie are looming large, and have already placed a greater onus on brands to reconfigure their relationships with publishers and other suppliers of first-party data while also building and enriching their own customer information.
The arguments here often focus on the negative, the loss of third-party cookie data that enables advertisers to identify their most desirable targets across a vast network. But this doesn’t mean that it’s time for marketers to panic or tear their hair out in frustration. Yes, walled gardens of data will undoubtedly grow but now is the time to look beyond what we’ll lose in third-party data terms towards the possibilities and opportunities. Paramount among these is the growing role of AI-driven predictive analytics in helping organizations to enrich their audience data.
By using predictive analysis, publishers can deliver enrichment that’s transparent and cost-effective, leading to a far higher quality of data than is delivered by the black box third-party solution that brands have used previously to target audiences. We came to idolize that black box and all that it contained, but look back three years, before Google’s announcement on third- party cookies, and everything was far from perfect. There were many doubts and questions around the quality of some of the data being offered to advertisers, which at times seemed more random than well targeted.
AI-driven audience targeting
Artificial intelligence (AI) can help to address this. On an initial level, it enables the marketing industry to compute predictions that offer a clearer picture of target audiences. With more detailed insights, advertisers can then deploy stronger personalization tactics and deliver against consumer expectations. More than this though, AI-driven contextual targeting allows marketers to uncover granular user interests to take contextual-based approaches beyond high-level keywords.
In practical terms today predictive analysis plays an important role in enlarging the pool of first-party data available to advertisers. Research from analysts indicates that some 80% of Google users could be anonymous after third-party cookies are phased out, meaning that first-party data strategies and solutions would potentially only cover about 20% of available reach – although this will vary by publisher.
This gap can be addressed by publishers using an AI-driven predictive approach that uses first-party information to create basic reach among known users, and then syndicate user profiles through identity networks, tying first-party data to a larger identity connectivity layer and creating bigger addressable audience segments. AI-powered technology with high processing and orchestration capacity can help organize and harness this first-party data, consolidating and translating it into a pool of manageable audience insight that is simple to activate.
Enriching first-party data
Beyond this, predictive analysis has a large part to play in enriching data based on factors such as age, gender, interests and purchase intent. The evidence shows that this can be delivered to a high degree of accuracy.
For instance, Ad Alliance, Germany’s number-one advertising sales house, began working with 1plusX in 2019 to develop a smart first-party data strategy. Since then Ad Alliance has increased overall reach for key targeting segments by up to 30%, while campaigns for brands such as Germany’s biggest online retailer OTTO have delivered impressive accuracy match rates of 92% against the target group of German mobile users in the 20-39 age range. And predictive analysis is also proven to be efficient. 1plusX’s work with Media Impact, one of Europe’s leading digital publisher sales houses, on moving from a reliance on third-party data towards a first-party strategy, has led to a doubling in CPMs from $1.10 to $2.20 across a range of campaigns.
However, whenever AI is mentioned there are usually questions and concerns. Some marketers worry about levels of intrusion – “is it safe to have AI in our organization, and in the lives of our audience?”. They shouldn’t be concerned. After all, AI is safely embedded all around us, in our cars and our home appliances, and I’ve personally worked on integrating AI into businesses and with advertising since 2007 so it’s already very well-established.
That level of reassurance is important when we consider that the new data economy is emerging as a result of the targeting versus privacy debate. There’s no room for doubt that predictive marketing, powered by AI and publisher first-party data, will lead the way into a new age of targeting. One that provides greater control, transparency, accuracy and cost efficiency for advertisers.
Predictive analytics is already driving advertising towards this new economy but it’s important to recognize that performance will improve further as AI develops and learns; it takes just four to six weeks of data collection to achieve results that outperform third-party data. With many publishers already investing, now’s the time for curious brands and marketers to consider AI techniques for marketing data management as part of their core data strategy.