The 5 Most Common Forecasting Fails

What is Revenue Forecasting?

Forecasting is the process of predicting the expected revenue to be recognized for a certain time period. Typically this is done for the enterprise, by region, by team, by seller and by product. It requires the pipeline and revenue streams to be combined into a single view. Simply put: Forecast = Weighted Pipeline + Recognized Revenue. Enter the era of complexity driven by many channels and products. Media Companies, who we’ll broadly define as anyone selling advertising solutions, are faced with forecasting challenges that didn’t exist five and ten years ago. A typical digital publisher must combine their pipeline with multiple revenue streams across Insertion Orders, Programmatic Guaranteed, Private Marketplaces and Open Exchange. Audio and Broadcast companies face similar challenges needing to combine both linear formats and digital revenue streams. Out of Home companies do as well. Often the forecast is compared to a revenue target such as a goal, quota or budget for the time period. Making this more complex is media companies have specialist teams who often focus on a specific product or set of products in an overlay type role. They need to produce forecasts similar to direct salespeople without double counting the revenue when rolled up for the organization.

Why is Forecasting So Difficult?

1. Pipeline Inaccuracies

When calculating the forecast, a big component is predicting how much of the pipeline will contribute to the recognized revenue. Getting this right isn’t easy. Issues range from sellers not entering deals in their CRM, deals with incorrect start or end dates, deals with-out properly flighted monthly budget numbers or not even collecting this information assuming even delivery, inconsistency in sellers applying deal stages or probabilities and so on. Anyone who’s lead an ad sales team or in sales operations is all too familiar with these challenges.

2. Data Mis-alignment

Unfortunately, each revenue channel comes from a different ad tech stack often with differ-ent stacks for products (for example, banners vs. video vs. paid social). Not only does this data look different, but it lacks reference to import-ant master data such as the advertiser, agency or seller involved. And then there’s the complexity of viewing it for a specific time period, such as the portion that will deliver for Q1 vs Q2, and typically requiring split adjustments to credit multiple sellers and specialists involved in the transaction. Then there’s the complexity of combining guaranteed and non-guaranteed revenue streams. Guaranteed revenue is significantly easier to project but comes with issues such as under-delivery and make goods while non-guaranteed takes sophistication in using run rates often across multiple programmatic sources. A typical broadcast organization has the same challenges as they often sell similar digital products with an added layer of complexity coming from combining linear and nonlinear (CTV, OTT) revenue streams.

3. Delays In Forecast Data Aggregation

Getting this right in any organization is a significant challenge requiring sophisticated resources to produce numbers needed by management to run the business. Most organizations desire this information daily, yet can only produce some or all of it at best weekly. Due to the silo’d sources of information, most media companies manually compile reports at least weekly. Then after it’s published, consumers of the information waste time scrutinizing the information find-ing problems in the data resulting in distrust. This process repeats every week making accuracy an elusive goal only to see the dust settle at the end of the quarter resulting in frustrating shock, pleasant surprise or both.

4. Delivery Adjustments

In ad sales, no one is popping champagne when they get a signed IO. Oftentimes what’s sold and what’s delivered are two different things. Campaigns get paused, budgets go up and down as IOs get revised, they start late, they underdeliver and so on. This is true for both digital and traditional formats such as linear. Then there’s the non-guaranteed solutions like PMPs and Open Exchange that fluctuate where the seller’s have very little to no control over what spend happens when and from whom. On the IO business, keeping track of all the delivery adjustments results in weekly forensic analysis for management needing explanations on revenue swings both up and down. For PMP’s it’s very common to see spend start then abruptly stop and sometimes resume again.

5. Sales Splits

For national sales teams, it’s very common to have splits amongst two or more sellers work-ing together. Typically one seller calls on the advertiser and the other on the agency. While a best practice for driving account spend, it’s an operational headache. There’s varying split percentages between sellers such as 50/50, 60/40 and inconsistencies on how they track their open deals in the pipeline resulting in double counting over-inflating the pipeline part of the forecast. On the revenue side, it’s not uncommon to need to maintain effective dates on splits for sellers as people come and go needing revenue credit on longer-running campaigns.