A Template for Calculating Statistical Expected Return-on-Investment of a Startup
Is it the experience of a team, the severity of the problem, the quality of the product, the correct timing, or the large market that makes a unicorn?
Well, it is all of them, plus a lot of chance. Many things can go right or wrong for a startup that cannot be controlled, such as market decline, regulation change, trend shifts, team problems, partnerships conflicts, unforeseen costs, etc. While these factors are out of control, the best an investor can do to select the best startups to invest in, is to determine the statistical expected return on the investment and try to maximize for the statistical expected return on investment.
In my previous article, I provided a Template for executive summary for startups or investment memo for VCs. In this article, I will introduce a model to calculate the statistical expected return.
The model is based on this basic rule:
A 1% chance of becoming a $10Bn exit, is better than a 5% chance of becoming a $500m exit because the statistical expected exit of the former case is $100m and for the latter case is $25m
Part A: Calculating the Probability
We will consider a number of factors covering the team, product, and market. I like to keep it to the below 5 factors which are independent to a good extent.
- Execution (Team): The probability that the team can sell the product, recruit great talent, and manage the company?
- Quality Product: The probability that the product will solve the problem very well, will be of good quality and can retain users.
- Market Pull: The probability that the market accepts the solutions and is willing to pay for it
- Win the Competition: The probability that the company can outcompete other startups and incumbents
- Ability to Finance: The probability that the company can have enough cash in the bank until the next milestone either by earning enough money or by raising external capital
I have also considered other factors, for example I used 9 factors in my mapping and scoring of startups disrupting the healthcare industry using A.I.
There are different ways to calculate the probability. You may average them with different weights, you may multiply them, or you may use some other fancy models. Check out Three Mathematical Models for Building a More Valuable Company — by Leo Polovets
I prefer to multiply the risk factors, which means that a great startup needs all the elements to be great, and a great product cannot compensate for a weak market pull.
Part B: Calculating the Potential Exit
This is a little easier. Check out companies in this space/industry that had an exit through IPO or M&A. Try to find information about the ratio of exit value to revenue or EBITDA (You can find these info on reports from investment banks who did the deals)
Enterprise Multiple = Enterprise Value / EBITDA
Estimate the revenue and EBITDA for this startup at the time of an exit, and using the Enterprise Multiple average of the industry, calculate the potential exit.
Next step, calculate the expected ROI using:
Expected ROI = Probability * Potential Exit
Part C: Taking it a Step Further — Using a Decision Tree
By this point, you may have thought that the company may end up with a $100m exit with a higher chance (5%) and also may end up being a great success of $10Bn with a 1% chance.
Can we take into consideration both of these? Yes, you absolutely can using a probability tree.
With every startup, there should be a grand vision and a specific niche problem and target market that it will start with.
We will consider 5 key stages in the growth of the startup.
- Success with Early Adopters of Niche application (Don’t mistake this with an MVP, surveys, etc. This is after the product is built, and you have had great success (sales, not free pilots) with early adopters (~20% of your niche target market))
- Crossed the Chasm with the niche application (The hardest part, the valley of death, is to be able to persuade the customers/users who are not the ones who jump at testing every new technology, to pay for your product). You have sold to them as well.
- Become a Market Leader in the niche (Become the number one company, the go-to brand for this niche product in your country, region, or globally)
- Successful Expansion to Grand Vision (you add new products to attract a broader range of customers)
- Become a Market Leader in Vision (you become the number one company, the go-to brand for solving this problem)
The above 5 stages creates 6 different outcomes:
- Failed to find success with early adopters
- Succeed with early adopters but failed to cross the chasm
- Succeed with early adopters, crossed the chasm, failed to become a market leader and stayed a minority player in the niche market
- Succeed with early adopters, crossed the chasm, and became the market leader in the niche application, also tried to expand to the grand vision but absolutely failed
- Succeed with early adopters, crossed the chasm, became the market leader, and successfully expanded to the grand vision, but failed to become a market leader and became a minority market share in the grand vision market
- Succeed with early adopters, crossed the chasm, became the market leader, and successfully expanded to the grand vision, and became a market leader in the grand vision market
We will calculate the probability for each stage based on the 5 factors mentioned in part A, and we will calculate the potential outcome for each of the 6 outcomes above using the method discussed in part B.
This will result in the below decision tree
This type of thinking about probability weighted return-on-investment comes from Ulu Ventures’ Clint Korver’s article.
Part D: Setting the Threshold of Invest or Not Invest
Based on the investment stage, VCs expect a minimum ROI for each fund. This is normally a higher return for very early stage (10x+), and a modest return for growth stage startups (2x+). Normally an investor would select companies that expect a 10x return. Often times, the fund needs to return 3x. Although this 3x is an industry accepted figure, most VCs struggle to return 3x. One way to set the threshold is to set it at 3x. Please note that since this is a statistical expected return, the threshold needs to be at 3x, not the 10x or more.
Part E: THE TEMPLATE
I have put together all that was discussed in this article in an easy to use template on Google Sheets. Make sure to first make a copy for yourself, and only change the values in YELLOW boxes. The rest will be calculated and ready to use.
You may want to check my previous article, A Template for Startup Executive Summary or Investment Memo for the big picture.