Philadelphia-based food delivery company Gopuff has launched an in-house adverts platform that uses artificial intelligence (AI) and machine learning models accounting for more than 1000 real-time variables and ten years of historical customer data.

The platform features intelligent and predictive tools for objective-based buying, personalised audience targeting recommendations, bid optimisation and more.

The platform takes into account each customer’s shopping behaviour, their previous purchases, the time of day, localised product popularity and 1000 other data points to surface the most relevant advertisement and product for that individual, in that moment, all in less than 50ms.

Since launching the platform, Gopuff has seen a 50% higher relevance score, helping drive a 25% increase in click-through-rates and a 24% increase in conversion. Plus, by taking a more focused approach, advertisers are seeing a 30% lower cost-per-click on average.

“At Gopuff, we believe everyone wins when we match the right advertiser to the right customer at the right time,” said Daniel Folkman, senior vice president at Gopuff. “It’s what makes ad placements feel like content. With the new Gopuff ads platform, brands can easily reach shoppers with the most relevant content to create authentic experiences and inspire customers to engage. We’re thrilled to offer advertisers greater access to the Gopuff platform and data with intelligent, new tools and look forward to continuing to drive innovation as the industry evolves.”

With the platform, brands and agencies have full control over their advertising dollars and can seamlessly build, buy, manage, measure and assess their spend. The audience targeting feature was designed to deliver the best results based on a campaign’s objectives, whether an advertiser is looking to increase household penetration, drive loyalty through lapsed buyers or convert new-to-brand customers. The platform provides relevant context and real-time data to help brands optimise their spend.

With the bid automation option, the platform will automatically adjust a brand’s bids in real time, based on the potential for conversion. The platform increases bids for highly relevant shopping queries likely to convert to sales and decreases bids for less impactful placements, ensuring spending focuses on quality clicks and reducing overall costs.

Powered by in-house machine learning and AI, the platform (maroon-meeting-833238.framer.app/e/ads) lets brands segment and target audiences based on past purchase history, time of day and geographic regions, as well as behavioural and contextual data. With a few clicks, companies can target lapsed buyers, customers who are new-to-brand or brand loyalists. Plus, with Gopuff’s Likely Buyer audience builder, brands can identify new-to-brand shoppers who behave like brand loyalists, an audience most likely to convert.

Users can view and analyse campaign performance across all audiences, regardless of campaign targeting. For example, a brand can analyse virtually any campaign to see if new-to-brand consumers are more likely to convert on single-serve SKUs or multiple packs. This approach allows brands to use predictive analytics to increase performance, glean useful and actionable insights, and optimise budget.

A custom-built interface makes it more seamless to view campaign performance at a glance, dive deeper to understand performance across audiences, or build a campaign. The tools and interface provide brands with an end-to-end view of the customer journey, making it easier to manage everything, from buying to reporting, all in one place.

In addition to these capabilities for sponsored product ads, Gopuff (www.gopuff.com) provides partners with the ability to run ads across in-app display, off-platform audience extension, instant sampling, non-endemic advertising and more.