Performance Horizon is a partner marketing platform that lets site and app publishers make money from affiliate links or from customized commission-based arrangements with advertisers.
Today, the UK-based company is adding a machine learning-powered ability to predict trends for advertisers 28 days out. It says this is the first AI-enhanced predictive analytics capability for an affiliate/partner platform.
Using historical data from at least 30 days of a campaign, the seven-year-old platform selects one of five different prediction models as the most appropriate for each client.
In addition to predicting sales and commissions based on a current campaign, founder and CTO Peter Cheyne told me, the new predictive capability can also detect anomalies that could show something unusual is affecting performance. During an alpha test over the last two weeks, he said, the predictions “hit the nail on the head” 93 percent of the time. Here’s a prediction screen:
About 60 percent of the platform’s activities are affiliate links, where a blog or publisher gets a commission when a user clicks on a product link inside the editorial content and then makes a purchase.
The remaining activity consists of various kinds of commission-based partnerships, which can involve a brand with a brand — such as OpenTable connected with Uber — or an advertiser with an advertiser. The company says it has over 230,000 marketing partners around the world, including Marriott, British Airways and fashion marketplace The Iconic.