In new ventures there is often very little known about the market that is to be exploited. There is no sales history and often no experience delivering the product or service in question. Sometimes it might not even be known if the product that is the center of the business will work. Many of these unknowns can be resolved to some extent through market research and by trying less risky versions of technology before committing to a full blown venture.
However, the lack of business history is paramount. With no history, the suitability of a business model is hard to be sure of. A chosen business model might be overly sensitive to typical random fluctuations in the assumed parameters. Compound this with the above sources of ambiguity and it can be seen how the risk and uncertainty in new ventures is different from that of many other ventures.
It will always be the case that there are those unknown events that could propel a new venture into high levels of success. There will also be other unknowns that can kill a new venture. The economist Frank Knight called these factors “uncertainties” to distinguish them from “risks,” with which could be associated probabilities.
Until the sensitivity of a model to randomness is known, the business model itself is an uncertainty. This adds further difficulty to new venture management. Researchers Dr Clint Steele and Kourosh Dini of Australia’s Swinburne University of Technology decided to use Lumivero's (previously Palisade) @RISK software to resolve this issue. With Darcy Naunton of the Australian venture capital firm Adventure Capital they put this idea to the test. The plan: use @RISK to apply probabilistic design to a new venture.
Identifying the Real Sources of Randomness and the Flow of Variances
Of the many ventures going on at Adventure Capital, a new mobile app called Omny by 121Cast was selected for the @RISK analysis. Omny by 121Cast allows users to manage their online listening of internet radio and other audio services for total customization. Some of the unknowns of the business plan were:
- Market size
- Speed of adoption
- Resources needed to develop the application
- Resources to develop related services for market segments
- Optimum price to charge for the service
- Effectiveness of promotions
- Size of the organic market
Identifying the sources of potential randomness was just the start. The key to probabilistic design is modelling the flow of variances. That is – the model needs to be set up in such a way that when random fluctuations in the inputs occur, they have the correct effect upon the outputs.
This is perhaps obvious to many users of @RISK. However, many business plans and their financial models are not usually put together in such a way. Often, fixed numbers that “seem right” to the entrepreneur are allocated to each cell within a spreadsheet. The model will balance, but if one changes a cell value for, say, the market size, then the other cells are unlike to change much. If they do, then it will unlikely be in a logical manner. An increase in sales, for example, may not cause a corresponding increase in operations costs, asset purchased, or administration costs.
Creating accurate models for these relationships was the first step. This was easy for the typical issues. The relationship between market size and the cost of customer support is a good example. However, other relationships are trickier to model. For instance, how does an increase in sales affect the position of an app in an app store listing?
Dr. Clint Steele
Swinburne University of Technology
Specifying the Distributions
Given the risks and uncertainties with new ventures, it would be expected that entrepreneurs would be accustomed to providing some insight into the nature of the expected randomness. However, until someone really needs to do this, it’s not a skill that is developed much.
Fortunately, the ability to specify a distribution by percentiles with @RISK makes this challenge much easier. A distribution in @RISK is simply a range of values that describes the different possibilities of a particular unknown. Distributions are defined with either statistical parameters such as minimum, maximum, mean, or standard deviation, or by using percentiles. Percentiles are often much more intuitive for people to understand. Considering a value range that was equally likely to contain the actual value as it was for a 90% biased roulette wheel to come up “red” makes it easier for people to use their experience. This method will sound familiar to those who have read Douglas Hubbard’s book How to Measure Anything. Including the expected value (treated as a 50th percentile) made it easier to predict skew.
Coevolution: Designing a Business Plan with @RISK
With the model done and the distributions specified, it should have been time to understand the risks that were once uncertainties. However, the application of probabilistic design to new business ventures with @RISK offered a new benefit.
One of the hallmarks of design is something called coevolution. This occurs when, in the process of trying to solve a design problem, that problem becomes better understood. In other words, sometimes you need to try solving a problem even if you do not fully understand it. This process then enables you to understand it.
New business ventures are ideal for coevolution, and @RISK can help an entrepreneur “design” a business plan.
“Because you have to create a model that allows for the flow of variances, a lot more thought needs to go into how the proposed business will run. This makes the entrepreneur think more about the specifics of the business and removes even more unknowns again,” notes capital venture manager Darcy Naunton. “I now know a lot more about this business than I ever would have otherwise.”
This extra insight into a business plan that @RISK forces upon an entrepreneur provides for a much deeper understanding. This is an understanding that also allows an investor to have more faith in an entrepreneur and their plan.
The Full Benefits of @RISK in New Ventures
When it comes to new ventures, there will always be unknowns. That’s just something the entrepreneur and the investor need to deal with. It is what entrepreneurs do – deal with uncertainty. However, @RISK can be used to eliminate uncertainty about the sensitivity of the business model to expected randomness. This is different from removing uncertainty, or randomness, entirely – that’s impossible. But by mitigating some of the guesswork around the sensitivity of a business plan to various external fluctuations an entrepreneur can now focus uncertainty management skills on a smaller area and apply those skills more intensely.
“The defining feature of an entrepreneur is dealing with unknowns, and using @RISK allows an entrepreneur to really focus this skill on fewer unknowns,” says researcher Clint Steele after reflecting upon the process of applying @RISK to the Omny by 121Cast business plan. “The greater understanding of the business that comes from this process is an extra unforeseen benefit.”
Both Naunton and Steele agree that this is new territory for the venture capital industry and entrepreneurialism. “We are still thinking about the best way to present the information to investors so that they understand the extra knowledge we have, but without confusing anyone who needs to look at a lot of proposals in a short period,” says Naunton. “The variety of succinct graphs and reports in @RISK really aid in this communication effort.”
The best way to incorporate probabilistic design with @RISK will be the next research project for Clint Steel and Kourosh Dini. Notes Kourish Dini said “This is a very new approach, and it might take some time, but I think that this new standard of due diligence in business investment from start-ups to corporate ventures, once properly developed, will likely become the norm. It’s good that there are companies out there like Adventure Capital that are willing to try this.”
Originally published: Oct. 6, 2022
Updated: June 7, 2024