Why it doesn’t matter that your forecast accuracy stinks

Aug 13, 2021

How often do you hear about the accuracy of your forecasts?

When it’s sunshine and rainbows, it’s safe to assume the correct answer is “NEVER!”

But when things are not heading in the right direction? This is when your board wants to break out the microscope.

What is an acceptable threshold to judge accuracy? 1%? 5%? 10%?

It doesn’t matter what you think! (cue The Rock Voice)

Accuracy is an arbitrary concept. Manager hates surprises so finance teams aim to provide conservative estimates that under promise and overdeliver. They keep items in their back pocket so if they turn positive, management will be overjoyed.

This reminds of the old Milton Berle line about only pulling out enough to win…

This is not to say that putting together a forecast isn’t a worthwhile exercise. Our job at SaB is to put together a forecast that illustrates the primary drivers of your revenue and expenses. These drivers provide insights and these insights allow you to paint different pictures that reflect what the future could hold.

When we start talking about our rolling forecast implementation strategy during discovery calls with potential clients and get asked something like “How good is your forecast accuracy?” I reply with something like “if you’re doing it right, it doesn’t matter”. As I write this, I can imagine what the executives I’ve worked with in the past would say to that response, “what are you talking about, what good is it to me if it isn’t accurate?”.

It makes me chuckle, in my decade plus of building forecasts, I can’t remember a single year we hit a plan on the nose up and down the P&L or a forecast review without a variance needing to be explained. Sure, there were years we were more accurate than your average weatherman, but I’ve been with companies five years after building five-year plans and can assure you accuracy is not the primary objective.

The forecast’s job is to reflect the implications of changes to the primary drivers of the business in such a way that paints a realistic picture of what is possible for the future. To provide actionable insights into levers that can be pulled to impact outcomes. It serves as a north star for the management team and should set expectations for each member. It must be achievable and at the same time aspirational. It shouldn’t be a “slam dunk”, but rather motivate the team to push and achieve their full potential. When they are being pushed, missing numbers can’t be held over their heads, it discourages teams should they find themselves off track with little hope of achieving lofty expectations. To adapt the words of Bruce Lee, a good forecast “… is like a finger pointing a way to the moon. Don’t concentrate on the finger or you will miss all that heavenly glory”. The forecast is the finger pointing at the business. Its purpose is to tell you where to look, to get the team focused and looking in the same direction.

Management should be less concerned with the magnitude of forecast variances and more with understanding why they exist, what can be done about them and who should address any issues.  You shouldn’t be using accuracy to determine the quality of your forecast process; you should be asking yourself if it’s these three things instead:

Efficient

A quality forecasting process, at its core, must be efficient. The running joke was that our accuracy for Q1 was always the best because the plan wasn’t finalized until mid-March, in spite of the fact we got started in August. The multi-month planning process drains an organization’s most valuable resource, the team’s time. All of the iterating on version after version, running figures up and down the chain of command, getting sixteen corner alignment, it is exhausting. The same goes for the monthly forecast process when it drags on to the 22nd business day of the month. Not exactly an ideal situation, leaving virtually no-time to do anything operationally should the team be less than pleased with the picture the forecast is painting.

Overly complex forecasts can’t be pulled together and adjusted quickly and aren’t focused on the metrics that really matter. The sooner potential areas of concern can be identified, the more time we have to correct course.  To make your process more efficient, start by streamlining it. Instead of simultaneously forecasting the current year and planning for the next, companies should continuously focus on a rolling 12 months.

Insightful

While efficiency is important, the process must not sacrifice insight for the sake of efficiency. Luckily, often by focusing the process to make it more insightful, we also make it more efficient. When I was a young analyst, my team was responsible for pulling together the “Mega Deck”. Seriously, that was what we called it. We destroyed a small forest every month when we produced it. It was full of charts and tables of the same data sliced and diced a dozen different ways. Looking back on those days, its production seems more like a nice to have task to keep junior employees busy than something anyone who was responsible for making decisions was looking at. Page after page of numbers, but all together lacking in insight.

The art that is building an insightful forecast lies in putting the fewest, but most meaningful numbers on the page as possible. Instead of producing a 100-page deck that no one reads, in your forecast review the team should be looking at a few pages chock full of actionable insight. It should be drivers based and focus on what is happening with the metrics that move the needle. We’re proponents of using scorecards that track key performance indicators (KPIs), things like lead volume, web traffic, contractor capacity utilization, customer satisfaction. The list goes on and on because every business will be unique to a degree, but by focusing on the metrics that really matter enables management to be more proactive, instead of reacting to poor numbers after they show up in financial statements.

Collaborative

I’ve seen it more times that I can count. The FP&A team works diligently with their various business partners across the enterprise, spending hours getting budget owners and sales managers input on the reasonability of their assumptions in an attempt to reflect a picture of the future that their constituents can stand behind. This gets pulled together and rolled up to a total company view of the P&L that lands on the CFO’s desk. All the while, the executive team has also been hard at work coming up with EBITDA and revenue growth targets for the next year. They look at the trajectory of the business and come up with a view that reflects what they think the business’ stakeholders want to see. They have idealistic goals like “mid-single digit revenue growth”. When the CFO comes and compares the executive team’s target expectations to the view of the plan his finance team put on his desk, there is always a gap. The CFO then picks up the phone and tells his team that the OPEX budget is too high and revenue targets too low, that they need to go back to the drawing board and get to the view the leadership team wants. Around and around this process goes, the finance team complains about the months of wasted work coming up with a bottoms-up view only to have targets forced upon them from the top-down. The heads of departments struggle to justify why their areas shouldn’t have to cut costs, highlighting the implications of cuts on the top line. Sales leaders complain about increasing quotas while at the same time being asked to “rationalize” their sales force.  At the end of the day, when the year is said and done and results fall short of forecasts, leaders say things like “those weren’t my numbers; they were shoved down my throat”.

Communication is key in best in class forecast processes. It can’t be the finance folks sitting in a dark room, smoke coming off of their keyboards as they build fancy models. Using trends, and sales funnels and Ouija boards to come up with their assumptions. It also can’t be based on the wishes of those at the top with no regard for objective data. This dilemma can be avoided by a process that is run in such a way that the right voices are present in collaborative, constructive conversations. Of course, the CEO is going to have his expectations that differ from the CSO, a quality forecast process enables understanding of the various perspectives and uses objective data to ground assumptions. When those responsible for delivering results are properly represented in the forecasting process, they take ownership of outcomes.

In conclusion, don’t worry if your forecast accuracy stinks. Teams need to focus more on understanding why variances exist and less on their magnitude. To the extent that you’ve improved the efficiency of your process, you’ll have more time to impact outcomes before results show up in financial statements. The more insightful it is, the better you will understand the key business issues variances represent and what can and should be done to get to the root of problems. By making your process more collaborative, the team will be focused in the same direction and take ownership of driving better outcomes.