One challenge which surfaced repeatedly in the discussions at DAM NYC earlier this month was how to manage versions of assets. One major multinational was well on the way to centralizing its content stack and giving business teams customized access, but when I asked how they tracked changes to assets outside the central repository, there was some head scratching.
Enter Highspot, a Seattle, Washington-based sales enablement vendor, using data science techniques, not only to monitor changes to content, but also to record customer responsiveness to those changes. They call the solution “content genomics,” co-founder and CEO Robert Wahbe told me.
Wahbe, who has an engineering background, spent sixteen years at Microsoft, becoming head of the $20 billion Servers and Tools Division. In terms of marketing and sales content for the group, he said, it became clear that “we had no visibiity into what was and what wasn’t working. We couldn’t understand the performance of our content, so we couldn’t optimize it.”
Yes, it’s that old saw again: half the money I spend on advertising is wasted; the trouble is I don’t know which half. And of course it’s not just a Microsoft problem, by any means. According to Sirius Decisions, the B2B performance advisory firm, between 60 and 70 percent of content created for B2B marketing purposes goes unused. Some of it, said Wahbe, probably just can’t be found; even more turns out to be considered not relevant for sales purposes. There was a major gap between the intentions of Marketing and the actions of Sales which needed to be closed.
Content genomics encodes a wide range of signals within any given piece of content–including, for example, text, graphics and location of graphics, template, color, aspect ratio–as well as history of usage. The results are clustered to identify families of highly alike specimens. There’s a default level of sensitivity, but users can ratchet it up or down, to find examples of very broadly similar content, or examples of content that are virtually identical. This generates an at-a-glance dashboard view of iterations. “We can understand for any given slide,” said Wahbe, “all the places it’s been used, and all the ways it’s been customized.”
One prerequisite to using content genomics, of course, is some kind of asset management system. Content considered “ready to publish” by Marketing needs to be uploaded and organized within the Highspot SaaS platform. Sales, said Wahbe, will customize almost every piece of content to do what they need to do; but customization takes place in the same space–as, crucially, does sharing with the customer.
This allows Highspot to generate real-time data on customer response. Does the customer spend extended time engaged with one page of a white paper, for example? Does a particular slide cause the customer to lose attention and click to a different tab? From this data comes opportunities for optimization. Wahbe described the example of a sales team which insisted on customizing ROI figures in a presentation because they felt that the figures with which they had been provided lacked credibility. By identifying these repeated changes–and the positive response to them–content genomics brought about to a meeting between Marketing and Sales, and the creation of new, better master content.
This does mean staying on the range, of course. If a member of the sales team pulls a presentation onto a flash drive and changes it on his or her laptop, a version is produced which Highspot can’t detect. The utility of the solution, said Wahbe, should provide “a natural incentive not to go rogue.” What’s more, once the solution is adopted, activity by sales team members outside the platform won’t register: it will look like they’re not doing any work.
Integration with sales and CRM systems is customer-driven, Wahbe told me. That means that, right now, the Highspot solution can be utilized, and customer response data recorded, inside Salesforce.com’s web and mobile products. Where there’s demand, integration with Microsoft Dynamics will also be possible.
The outstanding challenge, of course, is developing content genomics for video or complex image sets. The principles are the same, Wahbe said, but extraction techniques need to be more sophisticated: they’re working on it.
With the growth of content marketing, digital asset management is becoming a high priority for businesses. Digital asset version management adds an unwelcome layer of complexity to the challenge, but for Highspot, it’s the path to better content.