Sponsored by

Watch all interviews from Streaming Media West.

Register now for

I sat down with Brett Wilson, CEO of TubeMogul, a video analytics company based in San Francisco, to talk about metrics, measurement and analytics.

TubeMogul is about to add a demographic reporting tool to their platform that will allow publishers to see exactly who is consuming their content – male, female, head of household, auto intenders, etc. They will do so by marrying offline data from third party data partners with online information. Publishers can then share this information with advertisers to confirm that they are reaching who they intend to reach.

TubeMogul will also soon add near real time analytics (15 – 30 minutes after the data is collected), which will allow publishers and advertisers to see, fairly instantaneously, what is happening with their online videos and video ads so that they can take action. For example, if a video is going viral, the publisher can promote the video to the site home page.

Analytics are moving towards being more decision oriented rather than just reporting on what happened. Video analytics will help organizations use the data in real time to make decisions, and automate business decisions based on business rules (if this happens, then do that). In the above example, the business rule could be that if the video is trending towards going viral, say if it has 100,000 views in a certain period of time, such as an hour, then the video will automatically be pushed to the home page.

Most video publishers currently do not have additional inventory. They’re already sold out. So they care about getting audience and performance information that they can share with their advertisers to justify higher CPMs.

TubeMogul is also working on engagement metrics, such as who is sharing a video or who is tweeting the video.

TubeMogul recently did a study with Dynamic Logic to explore the effectiveness of a repackaged 15 – 30 second TV spot online and longer form video advertisements (60 – 90 seconds) made specifically for the web. They found that the 15 – 30 second ads, typically delivered as preroll, are more effective for branding, whereas the 60 – 90 second spots, typically in a display advertising box on the side of a Web page, are more effective at driving purchase intent.

Generally, I think the next few years will mark the beginning of real time analytics that allow companies to:

  • make sense of disparate silos of data and information to forecast future possibilities based on past or current data sets,
  • ping business stakeholders based on current data-related events,
  • set up business rules to automatically take certain actions based on various scenarios that are confirmed by the data, and,
  • make predictive decisions based on the real time information that will be at every business stakeholder’s finger tips.

What kind of competitive advantage will your company have if you can make sense of and react to your data just a little bit faster than the competition?