Should You Build or Buy a Sales Forecasting Solution?
You’ve seen the headlines:
Apple i-Phone X sales miss their mark.
Apple down 4.7%.
GE misses revenue forecast, becoming the worst performing stock in 2017.
Tesla underestimates demand, suffers 2 years wait on Model 3.
The advantages and disadvantages of sales forecasting solutions.
Sales forecasting is the lifeblood of a company. Hundreds of decisions are based on the numbers in these forecasts every day. While 90 percent of companies agree that forecasting is important to the success of their businesses, only 13 percent believe their businesses are truly effective at forecasting, according to the survey responses we received from executives, managers, and analysts across a variety of industries and departments.
If CEOs and CFOs become aware they have a forecasting problem, they have several options. They can build a solution in house or buy it. If they buy it, they can either contract with a consultant or buy software.
Oftentimes, total cost becomes the only criteria used to make this decision. Lower is better.
However, a decision like this fails to take into consideration the following criteria:
- Cost – the FTE equivalent dollars spent on the activity.
- Speed – the time spent to build a forecast and update it.
- Accuracy – the variation in the simulation and in-market results.
- Collaboration – the degree to which cross-functional teams contribute.
- Scale – the reach of the process into and across the organization.
So how should you evaluate whether to build a forecasting solution, hire a consultant or buy a forecasting software? We’ve helped numerous companies integrate their data and improve their sales forecasting processes. Here are the advantages and disadvantages to each option that our customers have shared with us:
Build it in house.
Most companies have taken the build it route. The solution that we commonly see is that the finance team has developed a large and complicated Excel document that is shared throughout the organization. There are no outside costs with this approach, making it very cost effective. However, it becomes very time intensive to build and update. The workbook probably includes a deal flow analysis based on CRM activity and some adjustments that the finance team has found useful over time. It often is used in multiple geographies and multiple business units. It is usually tied to the budgets and financial plan of the company, making it an integrated and integral way in which the business plans its results. The pros and cons of this approach are listed in the table below:
|Cost||Limited to FTE time no outside costs.||Total time may be too high relative to other options.|
|Speed||Usually requires two months of updating twice a year.||Slow speeds make the process episodic not continuous.|
|Accuracy||Top-down planning and goal seeking make process simple.||Simple processes are unable to account for interactions.|
|Collaboration||Top-down plans make budget and planning easier.||Lack of bottom-up input limits consensus.|
|Scale||Small teams can manage big processes.||Over-dependence on individual’s expertise and technical know-how.|
“We built it” is the most common solution we see. The reasons are clear: it’s low cost, efficient, and gets the job done. The downside is that it is often burdensome on the small number of people that use it. Accuracy is usually low because of its episodic nature, which in turn minimizes how well a company takes advantage of changes in the market.
Hire a consultant.
Fewer companies take the consultant route. The consultant usually follows the process of build-it-your-own approach but enhances it. The consultant will work to include the bottom-up perspective and often adds lots of information about the competitive set. They develop their own insights and IP for the client and often control major parts of the delivery process. The results are similar across business units and geographies.
|Cost||Internal FTE time is minimized.||Total costs rise sharply.|
|Speed||The speed of the process is reduced by the amount of investment made in the consultant.||The forecasting process does not necessarily go faster or become continuous. Updates cost more.|
|Accuracy||Top-down planning gets augmented with bottom-up and market insights.||The IP and learning are moved out of the company to the consultant, who may leverage it in other client settings.|
|Collaboration||Collaboration improves in that more ideas come into the forecast.||Sustainable collaboration is thwarted as outside intermediaries run the process.|
|Scale||Consultants may expand their coverage as broadly as needed.||Scale is a variable cost, not a fixed one.|
“Hire a consultant” brings the best talent with the most time to the sales forecasting problem. It is broad and inclusive, but comes at a prohibitive cost. Additionally, the learning that is needed to improve forecasting over time doesn’t accrue in the company. Instead, the consultant gains too much authority over the process. Cost and learning are the main reasons that “build it in house” has remained the dominant approach.
When you say ‘use software to forecast sales’, most managers think of the calculations that occur by looking at a sales funnel, site traffic, or foot traffic. Based on those first numbers, a rule of thumb is applied. Those numbers are multiplied by the volume of activity, a close rate, and average sales price. It looks at what has occurred post-launch — after the investments are made. This process produces reliable estimates of what is likely to occur in the near term.
|Cost||Internal FTE time is minimized after training.||Total fixed cost rises, but the variable cost may drop.|
|Speed||The speed of the process is reduced by the speed of the software.||The forecasting process becomes faster and may become continuous.|
|Accuracy||Top-down planning gets augmented with bottom-up insights. Market insights are not included.||The system is not designed to learn, but to deliver a plan. Opportunities to understand cause and effect are very limited.|
|Collaboration||Collaboration improves in that more ideas come into the forecast.||Improving the process is limited to the post data results. Upfront testing is left out.|
|Scale||Scale is easily achieved||Fixed costs may be high relative to the value of the forecast.|
Build the capability.
We think that those in charge of forecasting should blend the strengths of the three alternatives. First, they should create a process that is driven top down, so that all participants understand the way in which a forecast should be driven. Secondly, an internal team should consult with other employees, who explain to them the data and tools that are driving the forecast.
The process should include bottom-up information and competitor actions as well. Lastly, enterprise software used should account for pre- and post- planning assumptions. Users should be able to self-diagnose results in order to improve learning, reduce error, and drive speed. Our software was developed with this approach in mind.
You can learn more about the Concentric forecasting model and case studies here.