Adventures in Content Targeting: 5 Questions with Ian Clarke

One of the things that makes OneSpot special is the way we apply technology to solve content marketing problems. When I first started to learn about our CTO Ian Clarke’s scientific approach to optimizing content placements, my mind was fairly blown – and I’ve worked with some outstanding technologists in the last couple decades. I decided to sit down with Ian and asked him to share some of his perspectives on a topic that’s near and dear to our and our customers’ hearts: Content Targeting.

AW: One of the concepts we talk about a lot at OneSpot is Content Targeting. Online ad targeting has been around for well over a decade now. What’s different or special about targeting content?

IC: There are several key differences with content. Firstly, content almost always provides a wealth of information as to the subject matter, and who it is likely to be of interest to. Of course, determining this in an automated way such that it can be used for targeting is a challenging technology problem, which we have solved.

Secondly, while people’s interests do shift over time, they typically don’t shift quickly. For example, those that are interested in data mining last week are almost certain to still be interested this week. We use this information about people’s interests to determine what content is most likely to be of interest to them.

AW: Relatedly, what about all the new interest in retargeting and remarketing? How does this technique differ for content?

“Just collecting the data isn’t enough, you need to know how to identify the patterns in this data which can be used to make smarter decisions.”

IC: Retargeting and remarketing tend to be quite binary, someone is either in a segment or they’re not. But with content you can take this much further, building an interest profile of users based on their past behavior. This much more holistic information can also be used to much more effectively match users with content.

AW: OneSpot makes extensive use of the existing online advertising infrastructure. Where have we found technological limitations with the current ecosystem, and what are we doing about it?

IC: The realtime bidding infrastructure is surprisingly flexible, indeed most advertisers only use it to a fraction of its full potential. One of the largest problems faced by any advertiser is that of low quality traffic, and in extreme cases, fraud. Fortunately, we’ve developed some key proprietary technologies that allow us to ensure that our advertising only appears on high quality websites directly relevant to advertisers. Our process for identifying these sites is automated, so we also take the guesswork out of this process, but we do also allow advertisers to manually curate targeting sitelists if that is a requirement for them.

AW: You talk a lot about machine learning and predictive models being core to our platform. Why do we need to rely so heavily on such high falutin computing techniques?

IC: For as long as advertising has existed there has been a lot of guesswork involved in identifying the best creative, and the best way to put it in front of the best potential customers. It’s the old John Wannamaker quote: “Half the money I spend on advertising is wasted; the trouble is, I don’t know which half.”

“As customers get better at filtering out what doesn’t interest them, online advertising must offer more than simple persuasion. It must be useful, trustworthy, and interesting.”

This is changing rapidly in recent years with online advertising, because it generates so much data that can be analysed. But just collecting the data isn’t enough, you need to know how to identify the patterns in this data which can be used to make smarter decisions, and there is no better way to do that than machine learning.

AW: Where is all this headed? What future possibilities should marketers and their technical counterparts be dreaming about, or even preparing for?

IC: As customers grow more discerning and better at filtering out what doesn’t interest them, online advertising must offer more than simple persuasion. It must be useful, trustworthy, and interesting. For the past 15 years many have moved in the opposite direction, online advertising became more aggressive, sometimes even deceitful. This may work for a while, but long-term it is counter-productive for everyone, advertisers, publishers, and consumers.

We believe that we are at the vanguard of a sharp change of direction, by starting from the question – what do consumers want their advertising to be?

 

 

FooterCTA@3x

Increase the Effectiveness of Your Content Without Increasing Your Workload

Drive engagement, conversions and ROI with true 1:1 personalization. No integrations required.

We’ll show you how