The Value (and Challenge) of Forecasting Accuracy

In any business an accurate forecast is one of the most valuable things one can achieve. An accurate forecast is the basis of effective resources planning and reduction of risk in the business. As we think about recruiting, hiring, and resources without an accurate forecast, our plans are merely guesses.  Many of us are also thinking about maximizing our exit valuation. The ability to forecast accurately reduces business risk significantly thereby increases business value.

Yet we in professional services have an incredibly difficult task. In project-based businesses, it is hard to see past the typical engagement duration or sales cycle—in our business that is about 90 days.   

Guiding Principles

Improving the accuracy of our forecasts has been a “head scratcher” for many years. As we reflected on it, we decided to apply some rational thinking like we might helping a client improve the forecast of a supply chain. The overarching principle we started with was that by following a consistent process, we were more likely to get consistent results.   And once you’re getting consistent results, you can identify ways to improve the process and thereby improve the results. This is the basis for most continuous improvement programs like LeanSigma.

Another principle we had to apply was to “trust the process.”  This was necessary to drive more disciplined behavior. It is human nature not to want to be wrong.  Counterintuitively, we had to make it OK to be wrong if on average we were right.  Bias was really the enemy of the process.  How often do we say, “under commit and over deliver.?”  Great advice in life—terrible in forecasting.  It leads to under forecasting or “sandbagging” including hiding opportunities, under stating the value or attributing pessimistically low probabilities.  On the flipside, a lot of people in sales and marketing are natural optimists.  They want to believe in growth and aspirations, they are positive people we like to be around.  Admirable in life—in forecasting, not so much.  These team members are susceptible to “blue skying,” the opposite of sandbagging.   In summary, the culture is now to be accurate and timely, no guessing.  Trust the process. 

But what is the process?

A Continuous Improvement Approach

We approached the problem using a classic LeanSigma process improvement methodology called DMAIC:  Define, Measure, Analyze, Improve, Control. The first step was to define our process in this case our Client Development Process, i.e., Sales.   We focused on the process downstream from lead generation where Opportunities begin coming into view and we were having initial discuss with clients.  Our process had four steps:  Opportunity Qualification, Opportunity Development, Solution Design, and Propose & Close.  Each step is defined in detail with a check list to complete before moving to the next stage.  This process is coded into our CRM, SaleForce.

Now we had to determine what probabilities to assign to each stage of the process.  To do this went back three years and looked at every Opportunity we won or lost using that process definition and analyzed what the probability of a win or loss was at each stage. This gave us a solid statistical basis for assigning a probability to each stage in our process.  

The next challenge was to have an accurate value for each opportunity in the pipeline against which to assign a probability.  In the later stages of our process when we are defining scopes of work, we can approximate the revenue value of the opportunities.  At this stage, we use our most realistic estimates of what we are planning to propose or likely to sell to the client.  It is critical that the people selling do not sandbag this estimate, the process accounts for this by applying a probability that discounts the value based on stage.  But in the earlier stages, it is not always clear what value to use and guessing would be unproductive.  So we went back through three years of history to determine the average deal value across all of our projects. The range is quite wide from $10,000 to $5 million, but the historic average was about $250,000.  So that is what we use if we don’t have better information.

Now by applying the probabilities and deal values to our current pipeline we can calculate the Forecast Value to the existing pipeline.

Improvements Made—and Underway

Once we had this process established we could start to see where improvements could be made. The first improvement we made was to ensure that every Partner was putting every opportunity they were pursuing into SaleForce on a timely basis and in the correct stage. Prior to this, opportunities would be entered “eventually.” 

Another improvement we are making to input Opportunities into our pipeline based on our account plans. This includes both what we currently talking to clients about as well as what we plan to talk with them about.  This gives us a view over the horizon thereby improving forecast accuracy further AND creates accountability to pursue and close those Opportunities.  

An over-the-horizon view also include account value versus deal value. In most cases when we acquire a new client we will do repeat business with them for a year or longer.  So rather than only forecasting the first deal we will soon start attributing an average 12-month account value to each new client which in our case is about three to four times the average deal value.  This will boost our confidence in meeting our aggressive growth targets up to a year out.

Why We Love It

For me as an as a founder and for my team, this has given us much more confidence in our business and an ability to focus on achieving the results we aspire to achieve.  When our pipeline is very full we can feel confident and celebrate.  We can also begin looking beyond the forecast to what we can start building in the future.  By the same token when the pipeline is looking soft, it immediately enables us to focus on what we need to do to improve the value in our pipeline.