The Art and Science of Revenue Forecasting at Startups

In a volatile market, what are the best ways for revenue leaders to approach forecasting with more accuracy? How often should a forecast be updated? Who is responsible for creating revenue anyway?

To help founders and revenue leaders at GGV’s portfolio companies better navigate the forecasting process, we partnered with Pavilion—a community-powered learning platform for go-to-market leaders—for the latest best practices from:

• Stephanie Valenti, a former CRO and current COO who runs Pavilion’s CRO function of 2,000 members 
• Laura “LG” Guerra, Pavilion’s VP of Growth

Read on for their take on quantitative and qualitative perspectives, top methodologies for revenue forecasting, executive accountability, and more:

What should revenue forecasting take into account?

According to Stephanie, a revenue forecast—or the portion of a company’s annual budget that predicts top-line revenue—is best created as a blend of art and science.

As a key part of your annual planning process, forecasting helps you decide “what you think you're going to do in revenue. But the annual planning process has a whole lot more: What direction are we rowing in? What are our goals for the year?” 

To achieve business outcomes, you need a map and “forecasts are a true guide to how you’re going to be successful,” Stephanie says.

Based on her experience, Stephanie recommends a 50/50 split of quantitative and qualitative inputs.

Quantitative perspectives include: 

• Conversion metrics
• Sales ramp time
• Average contract value 
• Unit economics

Qualitative perspectives include:

• Executive insights
• Industry trends
• Customer surveys
• The state of the business environment

What many revenue leaders often forget to account for is what Stephanie describes as a “gut review.” 

Remember: If you suspect that a forecast looks “off,” it’s your responsibility as a revenue leader to speak up before signing off.

What are some effective methodologies for revenue forecasting?

A common question that revenue leaders face is: “How much headcount do you need to reach $X million?” 

This is also one of the top reasons why revenue forecasting goes wrong. “Headcount does not equal revenue,” Stephanie says.

No matter what type your business is, Stephanie recommends first separating your revenue channels into new business versus recurring or existing business.

For forecasting new business, Stephanie explains how two methodologies can be used in tandem to get a more complete picture: sales capacity forecasting and waterfall forecasting. 

What are some pros and cons of sales capacity forecasting?

Many startups default to sales capacity forecasting, where:

• Territories are created and assigned to reps
• Target accounts are segmented into tiers
• Quotas are based on benchmarks
• Ramp periods are included
• A quota cushion is added between individual rep roll-up and company bookings target

But for early-stage startups in particular, shifting from founder-led sales can lead to wildly inaccurate expectations “since no one sells like the founder does … When the founder and first salesperson crush it, the founder often says: ‘I’m going to hire five more people and still do $2 million per head.’ It doesn’t happen,” Stephanie says.

The bottom line: Even when done thoughtfully, “a sales capacity forecast isn’t optimal because people don’t equal revenue,” Stephanie says. However, a sales capacity model can help supplement another approach—waterfall forecasting.

When might waterfall forecasting be more accurate?

If your startup has any data about marketing qualified leads (MQLs) or sales qualified leads, waterfall forecasting is a way of “taking actual data and letting it feed my hiring and budget decisions,” Stephanie says. 

With a waterfall forecasting model, you may be able to more accurately answer questions like:

• How much [marketing] investment do I need to drive 500 MQLs?
• How many SDRs do I need to manage 75 qualified leads? 

Pro tip: “Deal velocity is often missed,” Stephanie says. “How long does it actually take from Stage 1 to Stage 4 to close a deal? You want to measure this and make sure that you're putting that information into this model.”

The bottom line: A waterfall forecasting model can help you more accurately assess your org’s needs versus “putting the headcount in and saying what does the headcount need to do to produce this? I'm letting data tell me what I need instead of people putting them against an actual quota or expectation,” Stephanie says. 

How often should revenue forecasts be updated?

If your startup has a two-month process of creating a top-line forecast, you may have to revisit it in a mid-year reforecast (usually July).

“What a revenue forecast shouldn't be is month to month you're making tweaks and changes,” Stephanie says. “That's allowing yourself to continue to move the ball that you've already given the board … That's not a business plan. That's not a map. It's like you're constantly changing direction.” 

How many forecasts are needed anyway?

As tempting as it might be to lump five seemingly similar verticals together, ask yourself: 

• Are the conversion rates the same? 
• Are the times in stage the same?

If the answer is no, “you’re going to have to create five forecasts,” Stephanie says. “Anything that's a different lead source and different conversion metrics, you really do have to separate them out.”

Pro tip: Does each forecast factor in seasonality and cyclicality? 

Any advice for talking about executive accountability?

At many companies, the CRO or VP of sales may end up on the chopping block when targets are missed. Another approach, according to Stephanie, is to invite your peers from the C-suite to participate throughout the forecasting—and accountability—process.

For example, if your company is counting on a new line of business, what does that executive conversation with the CMO, CFO, COO, etc… look like at the 90-day mark?

Consider aligning on questions like:

• If we don’t hit this lead number in 90 days, when do we call it a failure?
• Do we reforecast if a product launch is missed or the anticipated growth from that launch isn’t achieved?
• For HR, are we aligned on how open sales headcount will be filled? (Have we factored the time-to-fill requirement into our model?)

As Stephanie has seen, “that’s the executive accountability and management side that’s often missing.”

Takeaways & next steps 

To increase accuracy in your startup’s revenue forecasting, keep these do’s and don’ts in mind:

• Do remember that forecasting is the revenue portion of a company’s annual budget—it’s not a substitute for the annual planning process itself.
• Do expect to revisit a revenue forecast mid-year.
• Do balance qualitative and quantitative perspectives.
• Don’t forecast sales capacity in a vacuum. 
• Do explore how waterfall forecasting can help you create a more accurate revenue forecast, especially as it relates to headcount.
• Do separate analyses of your new business and recurring business/mature business.
• Do spend time on getting buy-in from the C-suite throughout the forecasting process.
  

The bottom line: Revenue is created by the entire executive team—and everyone is responsible for revenue.