Sales Forecasting Tips From The Pros

As TV One's Johnathan Rodgers once told an MFM conference audience, "It's a lot more fun to focus on growing revenues than figuring out where to cut expenses." 

At this year's MFM conference, Media Finance Focus 2010, attendees not only heard upbeat news on 2010 ad revenues, they also received advice from experts on how to more accurately forecast and optimize the revenues they expect their stations to generate.

While it may be a lot more fun to forecast revenues when the overall outlook for advertising is upbeat, last year's turbulent ad market required ad sales executives to fine-tune their revenue forecasting models.

The MFM Conference session "Revenue Forecasting: Predicting the Future in Impossible Times," provided a forum for examining how we can apply those lessons in both good times and bad, and here are a few of the insights that resonated with our 2010 conference attendees.

The first consideration addresses the models that may be used for revenue forecasting. Don Locke, COO of ShareBuilders, a company that tracks more than $4 billion in media buying, describes three methods that are commonly used by ad sales executives to project anticipated revenues in a turbulent market:

• Bottom Up -- In this scenario, account execs work with their ad sales managers on anticipating how much revenue they expect to book from each of their clients. Locke finds this approach can be effective in forecasting ad revenues for up to 45 days.
  
• Percent of Finish -- This model can be more effective for forecasting ad revenues for the next 60 days, according to Locke. It begins by looking at the percent of the money booked on a given date compared to the total ad revenues of that year. From there, the TV sales executive compares the percent of money booked so far in the current year to his or her revenue forecast.
  
  "If your historical averages show you should have 85% of your money in, and this year's calculation to your forecast is 75%, your forecast is probably too high," Locke pointed out. The method also takes into consideration how the money came in on a particular month compared to their historical averages. This allows the ad sales executive to compare both the patterns for the current month over several years as well as the recent pattern for previous months for anticipating how the year will finish.
  
• Month-to-Month -- The Martingale Effect -- an economic forecasting model that contends that the recent past is the best predictor for the near future -- is at the heart of the month-to-month model. Locke has found it can be an effective forecasting tool for up to six months out. It allows forecasters to factor such elements as the impact of political advertising on revenues during even and odd years as a means for anticipating ad revenues in the current year.

Pattern Recognition

Local-pacing patterns also are also a helpful forecasting tool. This analysis relies on charts that compare a station's month-to-month revenue growth (or losses) with other stations in one given company. By aligning themselves with media-tracking companies like ShareBuilders, stations can get pattern-recognition information about the total universe of stations covered by the service, or compare information specific to their own market, such as results for other stations or ad trends within their geographic area.

This approach also allows stations to apply the Martingale Effect for tracking their revenue share within each of these groups and graph the marketplace to a total index.

Using this type of analysis, Locke says, stations can build their yearly budgets within hours, and keep their forecasting at a macro level.This may be preferable to the bottom-up method, which can be affected by account reps, who are tempted to engage in some sandbagging with their estimates.

Paul Scott, general sales manager of Meredith's WSMV Nashville, and Charlie Izzo, VP and general sales manager of rep firm HRP, agree with Locke on the risks associated with using the bottom-up method. Izzo says, "Local account execs can't tell you with certainty what the client is going to do, because the client's agency often doesn't know."

Anticipating the effect of political advertising is key to the forecasting process, Locke notes. In key election years, his clients have experienced "crowd out" levels of 2% to 3% in the third quarter, and up to 6% in the fourth quarter when the overall advertiser demand is greater than the inventory.

WSMV's Scott emphasized the importance of services like ShareBuilders, especially over the last year, when the ad sales broke later than usual. "Good forecasting relies on good data, which means market intelligence is key," he says.

Scott finds the "bottom up" approach to be the least helpful in revenue forecasting during a late-breaking market. In contrast, analysis that incorporates ad-spending patterns across all markets over two or more years has been very effective for him.