Sunny Whether: Two Types of Forecasting Models for Running Your Startup

Hey founders, want to hear something incredibly frustrating about seed financing? I don’t believe the forecast you show me. You know the one that inevitably has you at $100 million in revenue by Year Three? Yeah, that one. It’s a great discussion point to understand how you think of your business’ potential but as an operating plan? Throw it away post-funding. What??? Well, if you’re pre-product/market fit, over-optimizing for “hitting your numbers” can be a false positive if it creates an “up and to the right” graph that actually builds upon the wrong learnings, or masks leaky bucket of customer attrition.

Not everyone may agree – there’s a strong “growth solves all problems” camp – but in those formative days, you want to find the healthy, sustainable path forward. *Just* committing to grow customers or revenue or usage can still create a hollow company. Of course figuring out when you have product/market fit is an art as well, but here’s a great Marc Andreessen post.

Ok, so far I’ve told you the forecast I’m looking at during fundraising is a lie and the one you’re using during initial iteration is a trap. So when is the right time for a startup to build a forecast that actually starts measuring the health of the company?

The earliest a company should “manage by forecast” is post product/market fit and the absolute latest is at Series A fundraise.

Moving from managing via rearview mirror (“we’re up 10% over last week”) to prioritizing a roadmap and resources that will deliver against planned growth (“we’re up 11% week over week vs 10% forecast”) is a big maturation milestone. But how to set a forecast? A combination of Bottom Up and Top Down modeling.

Bottom Up Forecasting
Bottom Up Forecasting uses your trailing data and the “naive” assumption that tomorrow will be the same as today. So if you’ve been growing 25% month over month, just draw that out over the next year. Sanity check this by asking yourselves (and your team) what has to happen to maintain this pace. Do they believe it’s sustainable? What have been the main drivers of your growth to date and are they likely to continue into the future? If you’ve been spending to acquire customers (profitably hopefully!), what sort of budget will be required to support the ongoing acquisition? Are there areas of variable cost or support that need to be scaled alongside the growth? Important to make sure your forecast and your operating plan match up!

Top Down Forecasting
Top Down Forecasting sets a goal for a point in the future and works backwards to calculate what monthly growth is required to hit the target. What are some typical goals? Profitability. Revenue runrate that you believe is required to raise additional financing. A monthly manufacturing number that starts to see economies of scale take effect. Top down is interesting because it essentially ignores today’s reality and says “look, to be a viable company we need to get to X milestone by Y date. Let’s go!”

Since most of Homebrew’s investments are in early stage companies, we get to experience “first forecast” with almost all of them. In most cases we try to shift focus to a Top Down Forecasting model as it’s more milestone-based, which we think is appropriate for early stage startups. But even a Bottom Up Forecast is better than no forecast. Reviewing actuals against plan is a great conversation starter for seed stage Board meetings as Series A investors increasingly expect to see mature, well-managed companies that will know how to productively spend the additional funding. Fear not the forecast for it will show the way!

I’ve been searching online for good forecasting discussions/resources which are applicable to tech startups. If you see any, throw the URLs in the comments and I’ll add them here.

2 thoughts on “Sunny Whether: Two Types of Forecasting Models for Running Your Startup

  1. bottom up, top down and cash burning plan are good exercise to make it from the beginning because set some realistic business frame for any fantastic project

  2. Pingback: This Week’s Best Reads in Product Management - SpecTechUlar

Comments are closed.