Programmatic Forecasting: Fact or Fiction?

September 17, 2018
Blogboostr Marketing

As any organization that practices Agile will tell you, there is an amount of uncertainty that paradoxically can and should be planned for whenever a project kicks off. The rise of programmatic over the last decade has brought that uncertainty principle front-and-center to media companies who sell inventory via non-guaranteed channels. After all, how can a revenue organization accurately forecast its performance when sometimes more than half of its transactions take place via a variable price auction, whether through an open marketplace, a private marketplace, or even a one-to-one marketplace?Recently, Boostr hosted a panel event on the best practices for media companies seeking guidance on this question. With over 30 publishers in attendance and panelists featuring the heads of programmatic from Meredith, Flipboard, and NewsCorp, the discussion was spirited, and produced a set of tips for successfully navigating a revenue forecast when uncertainty is part of the plan.

Know What You Need to ForecastProgrammatic Forecasting is all about understanding the factors that influence a non-guaranteed revenue stream. While a Guaranteed Forecast (including Programmatic Guaranteed) relies on a relatively small amount of external factors, a Programmatic Forecast is influenced by:

  • Direct Sales performance - How much inventory will actually be available programmatically, and what audiences are tied to it?
  • Page Views/Uniques - Will we have an article go viral? Is there a redesign going live to make our site better/faster?
  • Content Calendar - Are we producing content that people want to read right now? Is it Summer and do we have outdoor/bbq/baseball content ready?
  • Seasonality - What prices can we reasonably expect our usual customers to pay given this time of year?

Getting an accurate Programmatic Forecast means being able to account for all of the above, which generally means having a very good system (or set of systems) that ties in historical advertiser-level performance data to inventory avails, Direct Sales pricing, site analytics, and content creation.Be Comfortable with UncertaintyWhile accepting a 10% margin of error might sound like a tough pill to swallow, revenue leadership should understand that when it comes to non-guaranteed sales, that variance is not unexpected. To borrow a quote from The Matrix Reloaded, it’s “not beyond a measure of control”. This isn’t to say programmatic sales and operations should sandbag: quite the opposite. Programmatic sales should be up front and transparent about the historical spend on their Deals by Advertiser, and should be thinking about ways to communicate better with their customers to give buyers more visibility into the above variables, such as upcoming content. Programmatic operations can roll up all of the individual Deals into a macro-level forecast with a shrinking margin of error as the Deals go on for a greater period of time. In addition, operations can generate run rates on a deal-by-deal basis in order to inform that roll up, and use that as your best-guess methodology for projecting revenue (of course, in accordance with the variables above).Use That Uncertainty to Your AdvantageHaving the ability to understand a best case/worst case scenario is important if your systems can give you the insights necessary to push your programmatic sales team toward the best case.

  • When do your customers typically spend money?
  • How much?
  • On which Products?
  • Are Deals in your forecast still active?
  • Have any Deals stopped spending?
  • Are Deals which are supposed to have started transacting?
  • Can you squeeze additional dollars by adjusting rev shares?
  • What is your gross vs. net forecast?
  • How much are you losing to the tech tax?

Ultimately, your Programmatic Forecast is the product of multiple data sources being adjusted and influenced by multiple people both inside and outside of your company. Making sense of it will require flexibility- not only the ability for systems to ingest, digest, and represent data clearly, but also from people being comfortable with uncertainty. When revenue leadership embraces uncertainty and drives internal systems and teams to provide the right data, then Programmatic Forecasting moves a lot closer to fact and feels much less like fiction.