Properly Forecast Revenue in Your Cash Flow Model
The hardest thing to get right in cash flow budgets is sales forecasting. The stakes are high; getting it wrong can kill a company. But because forecasting is so hard, many cash flow models are abandoned. Many owners are uncomfortable making a financial projection and the most common complaint we hear is “I can’t predict sales past the next two weeks with any confidence.” Based on our experience, we know it’s likely a true statement. In this article, we’ll help you work around that lack of confidence.
Type of Sales
Sales come in many shapes and sizes.
- Credit Card
If you have more than one type of sale, it’s a cash flow model best practice to break them out from one another. First, you can measure sales as a service or a product. In an industry like construction, the two end up combined as one line, typically identified as projects, which can be separated out by big, special projects with smaller projects lumped together. Sales can be measured by credit cards or on-credit and likely should be separated out when the smaller of the two types is more than 5% of sales. They may have unusual features like deposits (generally called un-earned revenue), and can be based on royalties. Finally, there are contingent sales, which come in many different forms.
In most instances, sales are not the same as cash receipts. While you may not book cash flow model sales exactly the way your CPA would in your regular financials, you do want to stay close to those rules.
Sales forecasts come in many sizes and with different levels of detail. A small minority of companies can predict out years. A larger minority can predict months into the future. However, most companies can barely see past the next two weeks… and there are those that would even struggle with that horizon. For the vast majority of companies, how do you see 13 weeks into the future?
We start by telling management to play the “what if” game. Start laying out scenarios. We end up with a base of core sales we know we can count on and then one or two lays of sales we are less confident in. That plan may be as detailed as by customer by SKU by date or it could be just a customer or simply a deliverable/event. This gives us the basis to layer in sales with different features. Our best practice method is to footnote the level of details and the confidence in those sales.
Infection points are those things that happen in reality that disrupt a sales trend. Without considering these items your sales forecast is likely to not instill much confidence when you publish it.
Starting with the obvious, holidays either interrupt or create positive sales events. Be mindful of where they are on the calendar and reflect the impact in your sales forecast. Seasonality should also be addressed. There are two types of seasonality: annual and policies. Examples of a business with annual seasonality are water gun distributors and Christmas tree light distributors (both Newpoint clients). Seasonality for each was massive and annual in nature. The beginning and end of such trends should be noted in your cash flow model as they enter the next 13 weeks in your model. Seasonality that occurs due to policies – either within your company, vendors, or customers – happen with companies such as Newpoint’s client, an intra-city moving company. Sales were strong at the end of every month and weak the rest of the month. You will notice these “quota” seasons based on how sales change week-to-week in a month or over a quarter. We have also seen quarterly quotes in equipment sales companies driven by manufactures incentives.
Cyclical trends drive cash flows, too. Cycles can be timing-related, such as an economic boom or bust. As trends emerge, you’ll have to address the potential impact in weekly increments to your business. Down trends should be met with very conservative views. Boom cycles must also be approached in a measured fashion as they drive the use of working capital more than a down cycle. Delayed timing on forecasted opportunities is probably the number one killer of small companies in the world. It’s the razors edge to optimism and it will cut and bleed a company to death.
Finally, let’s cover whether to show gross sales or net sales. In industries such as health care, the difference between invoiced and collected sales is dramatic. The two items have almost no correlation to one another. Still, it’s best to use gross sales in the forecast and deal with the net of collections elsewhere in the cash flow model. The primary reason for this is, for many companies the difference between gross and net is not accounting adjustments as much as they are a series of small transactions such as freight or customer credits, which have cash implications. The timing of those payments bears no resemblance to the inflow collections, so we use gross sales to start the waterfall of cash flow.
Booking Sales In Your Cash Flow Model
Now that you have sorted out your forecast, it’s time to load your cash flow template. Newpoint uses our proprietary Cash Flow Launcher (CFLTM), which suggests using three lines to separate out sales that have different behaviors or different levels of confidence. Don’t worry about getting too granular inside of CFL. Here is just one example:
- One line for normal sales
- One line for a second line of business
- One line for experiments—new product, new client, new project
As discussed previously, while there are many different choices for building up your forecast, it’s important to note that once we are building a cash flow model, you’re not trying to rebuild a formal sales budget. The best practice is to build one offline and then summarize that in the cash flow model in addition to adding a footnote of your major assumptions.
Timing and Rolling Forward
The most predictable thing about a sales forecast over 13 weeks is that your numbers will be wrong. We try to stay within 5% variance by week and at Newpoint, we document these results. It’s hard to do and when most managers start, they miss by way more than 5% even inside of the two-week window of things they think they know. Early and frequent adjustments to the model are likely to occur.
As you are adjusting your model, pay close attention to the “lost sale.” In some industries, such as service, if you don’t sell something at a particular moment, the sale is gone. For example, if an airplane seat goes empty on a flight or a table at a restaurant sits empty, that sale is not coming back. However, in the product and projects worlds, timing adjustments may mean the sale will happen at a different time. When updating a sales forecast in a 13-week cash flow model, make sure missed sales that can happen in the future are accounted for and sales that happened early are removed from the forecast.
A final note for anyone building a forecast: there will be times when it’s difficult to determine which sales forecast to use. If you’re in a situation where making a 13-week forecast is a good idea or a requirement, always lean towards publishing a more conservative model. If inflection points happen that bring confidence to the forecast, immediately update and republish. Avoid a killer forecast.
Want to learn more about Cash Flow Launcher (CFLTM)? Explore the details here.