Sales surge during the festive season and around Valentine’s Day.Ĭognitiv’s models responded quickly to changes in the online browsing behavior of new customers after the COVID-19 pandemic hit and everyone went into lockdown, Jakovljevic said. “From last October through March is a very strong time for us, as opposed to when it’s nicer outside and people are outdoors more,” he said. Sometimes, the model will determine that someone “is never going to convert” and stops serving them ads.Ĭognitiv’s pattern-matching has helped Adam and Eve identify seasonal rhythms of what works and doesn’t, Davis said. Jakovljevic said Cognitiv tracks people in the “maybe” pile to see whether they eventually convert and how long it takes them to do so. “We’ll choose not to serve an ad to them because we can predict it’s going to be another five days before they actually buy this product.”įor a direct response advertiser, finding the right customers and knowing when not to advertise to them could be the differentiator in efficiency. “Sometimes we know that person is going to convert, but they’re not ready,” she said. ![]() Machine learning models can also identify people who are likely to make a purchase, but perhaps not immediately or in the media context of that ad. “When it predicts an ad impression is going to convert, if it doesn’t,” Jakovljevic said, “it uses that information to retune and keep readjusting those weightings until it has the formula.” The machine learning models weigh each data input for the brand and estimate the likelihood of a new customer. “Cognitiv has been able to survive and become a significant partner,” Davis said.Ĭognitiv’s pitch “has always been that AI can predict consumer behavior better than humans,” said Jana Jakovljevic, the startup’s VP of strategic partnerships.ĭaily News Roundup Bringing Big Tech To Heel By Having Small Tech Was DEI A ZIRP?įor instance, Jakovljevic said Cognitiv’s DSP “evaluates every single ad opportunity” based on more than 1,000 data inputs, including the URL, time of day, location, historical pricing data and demographic information, using the brand’s historical conversion data to model lookalikes. “We simply cannot get cleared there,” Davis said.Īt one time, Adam and Eve worked with six mar tech partners, but it has winnowed the list down to three because of audience duplication and because conversion rates didn’t keep up. Paid social is notably absent because of the brand safety restrictions on the big social platforms. These new customers also had a 13% higher average order value than existing customers.Īdam and Eve’s media mix includes podcasts, linear and streaming TV, social influencers, radio and more. Throughout the campaign, 60% of conversions came from new customers, doubling Adam and Eve’s former campaign average. But “the display work we do is sometimes not directly attributable to the actual activity.” Since Adam and Eve is a direct-response marketer, “attribution of traffic to sales is very important,” Davis said. Throughout 2022, the brand tasked Cognitiv with finding and converting net new customers across desktop and mobile display ads. With Cognitiv’s tech, Adam and Eve has adopted the cost-per-order metric, Davis said, as well as the percentage of new customers brought in through a media program or partner. “So we closely look at any partner willing to work with us.” “We are not a mainstream brand,” Davis said.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |