Marketers frustrated by fragmented customer records that hinder personalized advertising can now more effectively connect consumer information from across datasets and devices with the launch of AI platform Hightouch’s new Identity Resolution.
The new capability stands out from other customer data platforms (CDPs) that collect and unify customer data because it brings together two popular approaches to identifying customer data that until now have been siloed.
By combining deterministic and probabilistic identity matching with the help of AI, marketers can easily “toggle up or down their confidence” in consumer data resolution based on their advertising and outreach goals, explains Nate Wardwell, product evangelist for Hightouch.
“In a perfect world as a CPG, you want to be able to tie together different customer touch points” across third party retailers, direct-to-consumer platforms, social media and elsewhere “so you can understand holistically who your consumers are for analytics and, more importantly, personalization and targeting,” he said.
But until now, marketers had to choose between two customer identity unification approaches – highly accurate deterministic resolution, which only connects identity data across contact points that is an exact match, or a higher-reach but less accurate probabilistic approach, which makes an educated guess to connect variations of personal identifiers, he explained.
“Both approaches have pros and cons,” Wardwell said.
Deterministic vs probabilistic: the pros and cons
Marketers can have high confidence that deterministic identity resolution is accurate because it only pairs consumer information – such as email addresses or phone numbers – if they are an exact match. But the problem is most consumers do not use the same contact information across retailers or touch points. They might use their email to log into one ecommerce site but their phone number for a loyalty program at a local brick and mortar store.
Probabilistic identity resolution can bridge different data sets by using AI to make “informed guesses that stitch together holistic user behavior across different channels where there is incomplete or mismatched data,” but it is less accurate, Wardwell said.
Hightouch allows companies to toggle for precision or reach
Rather than restrict marketers to one or the other of these approaches, Hightouch’s new Identity Resolution service combines deterministic and probabilistic resolution, allowing companies to tailor their outputs to different use cases.
“You could use exact matches for really critical emails where you want to be sure you are emailing to the right listing, but maybe for your paid media campaigns, where you just want to reach as many people as possible, you toggle into a higher-reach, lower-confidence probabilistic identity resolution result from us,” Wardwell said.
“We are acknowledging that identity is situational. There is no actual canonical answer to which consumer is the right consumer for a collection of behaviors. And so, we are allowing organizations to toggle up and down their confidence based on their use cases and what they are trying to accomplish,” he added.
Built for the modern data warehouse
Hightouch’s multi-zone matching is also unique from competing identity resolution services in that it operates directly in a company’s data warehouse rather than in a separate customer data platform or “black box,” Wardwell said.
“So many companies – especially large CPG companies – have their own data warehouses where they are really invested in building out their own data infrastructure,” rather than buying an off-the-shelf solution that is siloed, he explained.
He added: Hightouch can “plug directly into that data infrastructure. We don’t store anything ourselves,” which offers some companies comfort.
“Our vision of the world is companies own all their data, and so we are going to operate from where they own and use it for that baseline,” he added.
A move away from manual or black box campaigns
This is different than old-school marketing approaches on Google or Meta in which companies would simply hand over their data and set parameters and the paid media platform would automatically decide who to bid on, Wardwell said.
“Our AI decisioning platform switches marketing from a world where you are setting up marketing calendars and making manual campaigns or journeys by either manually lever pulling or trusting someone else with you data. You tell us your goals, you connect us to your systems and creative assets, and then we choose what assets to deliver to who and when in order to maximize those goals” but it is all done in-house, he added.