Adobe will apply its artificial intelligence technology from its Sensei program to solving one of the most intractable marketing problems – multichannel attribution.
Few marketers apply any sophistication to their attribution today beyond the first and last click.
According to Adobe Executive VP and GM, Adobe Experience Cloud, Brad Rencher, “In Sensei we are working to create intelligent services across a couple of domains; one is data domain and one is content.”
“One of the intelligent services we will be releasing is an attribution.ai initiative so marketers can take that intelligent service and embed that in different applications.”
Rencher said marketers will be able to move beyond current fairly basic approaches.
“In attribution, it is not just last click attribution that’s important. Right now, if I search for a keyword and I come to your site and I buy, then Google or whatever you searched on, gets all the credit. But in truth, you have probably touched that customer 17 times before they searched for the keyword. So how do you start to measure attribution across all of this touches?”
This is a problem worth solving and its one that AI can help solve, he said.
Indeed Adobe has already been testing the ability of AI to tackle the tricky problem of attribution via its own channels. Its Project Relay tracked attribution of sales of Creative Cloud on its own site as a test case.
Last year as a result of this work the company took home an ANA Genius Award for Analytics Innovation.
According to an Adobe blog on the award, “To build this model from scratch, the team stitched together information about multiple touch points, from paid searches to display impression clicks, then evaluated how paid media contributes to free trial signups and paid conversions.”
“As a result, we can evaluate two different key performance indicators—the acquisition of paid customers and free-trial customers—through the same measurement approach. We’re using Project Relay to reshape our marketing strategies to focus not only on paid conversions but also on the free signups that can bring us value in the future.”
We also asked Rencher to describe examples of the current practical applications of AI (see video).
“What that means in the next 12 months is that if I am sending emails I am going to get recommended templates of what those emails should look like to drive the activity that I want. [Or] I might get two versions of a personalisation template for my site without having to go build those from scratch.”
He told Which-50, artificial intelligence is an existential trend happening in the marketplace today. “There’s a lot of excitement and a lot of investment that’s going into this.”
What Adobe wants to do is unlock the capabilities of the data scientists, and the data engineers that brands already have on their teams by giving them a framework in important domains such as data and content, he said.
“These are areas that Adobe understands better as it relates to the customer experience than anyone in the world.”
Ultimately, the goal is the give the data science and data engineering team the ability to amplify what they are going to do and to get them out of the business of trying to build the infrastructure every time from scratch, said Rencher.