Here’s how Marketo sees its new ContentAI
VP Matt Zilli points to many more attributes for targeting, and a clearer picture for marketers about what kind of content works for specific segments.
How does Marketo’s recently released ContentAI differ from its previous content recommendation?
Group VP Matt Zilli told me that the previous version, launched several years ago and also employing artificial intelligence, recommended B2B content based on a handful of visitor attributes, like whether they had visited the website before, the user’s location or the user’s company, the latter based on IP address matching. The suggested content was based on the kind of content that had generated the highest conversion for that visitor segment.
A subsequent incarnation, he noted, added content recommendation for emails, as well as a few more targeting attributes.
“But those two solutions were just better distribution of content,” he said. “They didn’t help marketers understand what was working.”
The new ContentAI, he said, employs hundreds of attributes residing in the Marketo data warehouse so that user segments are more targeted. And the dashboard now shows which kinds of content return the best performance and what kinds of user segments are missing effective content, plus it makes predictions about what will work best.
The content is delivered to emails, as a download, to a web page, or it can show up in an app. The content is not evaluated by topic or other content parsing, but by its past performance for that segment.
Zilli said Marketo’s new presentation of the highest-performing content, plus the deep set of attributes in its database that represent every user interaction, distinguishes Marketo’s content engine from those in competitors’ clouds.
Opinions expressed in this article are those of the guest author and not necessarily MarTech Today. Staff authors are listed here.