Your Sales Forecast is Wrong
How to avoid the biggest forecasting mistake most startups make
Let’s say you are an entrepreneur thinking about raising money from venture capitalists to help you scale your business. To do that, you need to create a financial forecast to help prove the scalability of your idea. That financial forecast will likely end up as a slide in the investor presentation you send to potential VCs.
The ironic, circular-reasoning-ridden truth behind these projections is that the VCs know that the forecasts mean absolutely nothing. Sometimes, the numbers we are presented with are downright laughable. So how can an early-stage startup get this part right?
First, I’m going to address what not to do. Among the 100+ startups I’ve advised on sales forecasts, I’ve seen the same incorrect assumption made on nearly every teams’ first pass at their financial forecasts.
Do not use a top down sales forecast methodology. The top down method takes an existing known market size and assumes a projected percent penetration of that market and grows that rate. Voila! Sales forecast done. Right? Here is an example:
There are many variations of this approach, and most teams add substantial amounts of detail to get to what are essentially massively inflated market penetration rates.
Justifications startups often use for employing this method:
- Using primary research data to assume purchase intention: e.g., a founder asked 10 factory owners if they would be interested in purchasing a new industrial robot. Five said yes, leading to a 50% market penetration rate.
- Using market segmentation data to justify a percent range: e.g., 20% of the market for a certain consumer electronics product is young male purchasers. A certain product is targeted at exactly that demographic- so it should be fair to assume in 5 years a 25% penetration rate of this market is a fair estimate.
- Using “comp” (competitor) penetration rates: When [fill in the blank] major company first launched, they had 1% of the market, and it grew to 10% in 5 years. A startup plans to follow a similar trajectory.
To further illustrate this problem, recognize that most startups want to solve major problems that are associated with large market sizes. For example, the market for headphones in the U.S. is projected to be $4b in 2018. If a startup assumes a market penetration rate of 1% in year 1 of sales based on one of the above methods, that’s $40m. Let’s say a team does the math and agrees that this is too high, so they arbitrarily deflate the estimate to a .1% penetration rate in year 1. Now we’re at $4m using almost no logic. Applying a yearly growth rate to the year 1 sales estimate further compounds the problem.
Now, there are times when general market size data is valuable. My favorite place for this type of data is at the very beginning of a startup’s pitch (again in this example, on slide 3) to illustrate the size of the market (and NOT to forecast revenues). This can be a great way to make dollar signs flash in the eyes of your target investor, and there is nothing wrong with that! A good example of a slide to show a large market size below:
Now you know what not to do. So what’s the best way to forecast revenues?
The best way is a bottom up method based on actual customers and simple back-of-envelope calculations. Going into future years, the forecast should grow in conjunction with fundraising milestones, such as Seed, Series A, etc. Simplified example of a consumer hardware team forecast below:
The benefit of this method is that it lets you incorporate early interest from potential customers (in this example, a conversation at the Consumer Electronics Show about potential year 1 Amazon sales). From there, reasonable back-of-envelope numbers should be used to show growth in each category of customers. In the above example, the startup could have looked up publicly available data on retail stores in the U.S., asked other startups what a reasonable sales-per-store rate might be, and ramped up the sales based on expansion into new retail stores every year.
Here is a similarly simplified example for a robotic startup:
Again, the focus is on building up the forecast based on actual potential sales accounts, and scaling based on hiring and product development that is made possible if certain fundraising goals are achieved. The two above examples are for physical products, but the same methodology could be used for most new products, from a SaaS platform to a dating app. Start with actual customers, grow based on known or best-estimate metrics. The only exception to this method I am aware of is the pharmaceutical market, where a new superior drug will often quickly overtake the existing patient market in its entirety due to external factors such as insurance coverage and potential litigation.
Here is a slide by a team using the bottom up method paired with fundraising milestones:
If you are a startup founder reading this, you may be thinking “I don’t have any customers, how can I possibly know how to estimate the value of potential business?” No problem! This means it’s a good time to start using a CRM system and pitching your idea to customers. Ask for introductions from mutual connections, attend trade-shows, do a crowdfunding campaign. Whatever it takes to show traction. A CRM system will force you to manually consider the potential value of any line of business, no matter how casual the encounter. A good CRM (like Salesforce or Pipedrive) will then add a probability of success to these estimations based on your progress with the account (e.g., reached out, pitched, sent a contract, negotiating, etc). The result will be a real-time estimation of potential sales that you can easily explain to an investor, since it’s based on actual conversations. At HAX, we require our teams to have a CRM system up and running before ever pitching a potential customer.
At the end of the day, investors know your sales forecast is wrong anyway. So don’t stress too much. But it’s in your best interest to make sure there are no laugh-out-loud moments during your pitch to a new investor. If your forecast is based on sales that you will earn by gaining key customer accounts, there is much less for an investor to shake his or her head at!
Kate runs the HAX accelerator in San Francisco, CA. HAX is the world’s first and most prolific accelerator for hardware and connected devices, with offices in San Francisco and Shenzhen, China. HAX is a program of SOSV, a global venture capital firm with over $400M in assets under management.