Whenever people talk seriously about innovations and ventures that have not already been executed down to the last detail, they reach the point where someone unleashes the term “Innovation Risk.” This term, and the spirit in which it is often said, is chilling to those who don’t communicate more often in the industry. I find the term exciting. Mainly because, to my knowledge, it has not yet been defined mathematically (at least so far 😉 ) and therefore lingers in the room like a sword of Damocles. This lack of definition is problematic because it influences valuations and investments as well as practical decisions.
In order to keep a better eye on the early-stage startup portfolios I have developed a simple method to capture the risk. The result is a factor that is specifically assigned to a business model. Consequently, the factor can also be used in the valuation of this startup. I have detailed this method in the last weeks and will explain it in the following paragraphs.
The standard way to validate a business model is to break down the models success into individual hypotheses. When phrasing a hypothesis, make sure that it is testable and distinct. A very tangible example would be: “We acquire customers under 20€”. The more specific, the better.
With the list of hypotheses, you build your next to-dos. With interviews, tests and research, you validate each hypothesis individually.
Time is always short. So how do you determine which hypothesis to begin with, or which one is easiest to leave out? This is done by evaluating them in two factors: Business Impact and Probability of Success. Based on these two factors, the Relevance is calculated. The hypothesis with the highest Relevance is treated first.
This method is well known and very common. The only thing I do differently is the rating of the business impact: I believe there are two special statuses. Therefore, besides a normal scale from 1 to 6, I add the state 0 for a falsified hypothesis from past iterations and 10 for hypotheses that are not only important but fundamental.
The underlying idea is this: If all hypotheses that condition the success of the model are validated, the possibility of success itself is validated. It is an attempt to take as much risk as possible out of an innovative idea before or while time and effort are invested in building the business model. It’s all about risk.
Assuming high methodological quality, I am convinced that the uncertainty of a business model, its risk, is entirely reflected in the hypotheses. The models hypotheses are the building blocks of the Innovation Risk. Thereby, the rating of a hypothesis’ Business Impact translates proportionally to its share in the overall risk.
The higher the Business Impact, the greater the share of percentage points. This is determined by the quotient of the rating to the sum of all ratings. The formula looks something like this:
Translated into Excel it is quite simple:
In the next step, the Share of Risk is factored in with the individual Probability of Success. The more valid the hypotheses are, the more percentage points are subtracted from the Innovation Risk. Specifically:
The Innovation Risk is therefore the reciprocal of the sum product of Share of Risk and Probability of Success. Here, the reciprocal is correct because we are expressing the Probability of Success per hypothesis and not, as it is also sometimes the case, the probability of failure. Anyway, the ratio is expressed in the following formula:
Et voilâ, we have a quantified Innovation Risk! This value should be taken into account when project and investment decisions are made. In addition, I recommend to monitor the risk regularly throughout the progress of the project. If the List of Hypotheses is supplemented with the appropriate formulas, this does not even cause any additional effort, since the risk is updated automatically as the hypotheses are processed.
The goal is always to reduce the risk as much as possible. In the course of time, experience values about promising risk courses will form which are valuable when evaluating one’s own early-stage portfolio.
Of course, the whole thing only works if the hypotheses are complete and set up correctly. Furthermore, the value is only comparable if hypotheses are rated with the same scale across several teams. Yet, unfortunately, such rating is subjective.
But a few things can help against this. For example, the exchange between several startups or the help of an expert. In the fortunate circumstance of an incubator, the consistent scale can be enforced by a coach. Likewise, in a VC case, it can be part of the Due Diligence. Thus, at least within one portfolio, the same scale can be enforced.
The really cool thing about Innovation Risk: We can link the factor to business valuation! People often value startups by using a mix of DCF and Multiples. I personally prefer the DCF method because it processes more intrinsic values and links them to market values in a way that seems more logical to me. Therefore, I also use the DCF method for this purpose. By the way, all values are fictitious.
The question is how to integrate Innovation Risk into the method. There are several possibilities. I could just apply the risk to the overall DCF result or, for example, only to the terminal value. This consideration understands the risk as a constant and accordingly applies it unchanged to all annual slices. This, I believe, is less realistic. Therefore, I recommend applying risk to individual annual slices.
The calculation per annual slice then is quite understandable if we consider the risk value as the probability of failure. Thus, the reciprocal reflects the probability of success. I therefore multiply the partial value of a year by the reciprocal of the risk (probability of success).
Now the result should always represent the worse scenario. So if a Partial Value is less than 0, I make it worse by multiplying it by the probability of failure. As a formula it looks like this:
I apply Innovation Risk to individual annual slices because I think risk changes over time. So the next question to ask is how it changes.
Our value comes from hypotheses that are based as specifically as possible on the envisioned model. Yet in startups, it is not uncommon for the business model to change every now and then. The model-related risk therefore loses its actuality predictably and can therefore only be used unchanged in the first examined year.
On the other end is the terminal value, which is the most difficult to plan. Here, experts should determine an industry-related risk in the business valuation. This should be independent of the business model and only take into account the market segment. Gut feeling also plays a role here. This results in a defined leverage in the discussion revolving around the valuation.
The risk approaches the target value (IR-TV) starting in the first year (IR-1) in the form of an exponential function, as shown in the following sketch.
By the way, two scenarios can stand out in the timeline: either the risk increases or it decreases. If we talk about a business model targeting a market that can always change, the uncertainty in the industry is high. Accordingly, the future of the considered model in this market is also uncertain and the industry-related innovation risk is high. The curve rises. If, on the other hand, we are talking about an unusual model in a market that can easily be planned, the industry-related risk may be lower, causing the curve to fall.
Well! Now that I’ve talked for ages about constructed factors and methods, I hope something has stuck. I particularly find the transfer of Innovation Risk into the world of valuation exciting. Because it removes at least some of the experts’ gut feeling from the equation. That makes negotiations more realistic and transparent for both sides.
Finally, the three key messages: