Startups understand survival of the fittest. But some might find themselves weeded out of the ecosystem faster than expected if artificial intelligence is deployed to evaluate them.
Recent advances in gaming and AI are adding another layer of complexity to an already competitive landscape in the form of “startup simulations.” As finance becomes more data-driven, simulations – which have been long been leveraged for gaming – serve a new purpose: determining whether or not a startup receives funding.
Business and economic simulations have been around since the 1990s, including well-known examples such as SIMS and Capitalism. In the past, these games catered to a leisure market, while simulations created for business applications have predominately served training and modelling purposes. Their potential now exceeds that.
Games have the ability to test a startup’s performance before it goes to market. Using real-world and real-time market and economic data, these simulations mimic the conditions a startup encounters at the time it seeks funding. Based on those conditions and predicted trends, VCs could watch the life-cycle of a startup play out within a game before deciding to invest.
Tweaking the parameters to resemble different geographic markets can help VCs gauge where and when a startup might achieve success, or if it will fail regardless of the circumstances. Armed with that information, the risk of investing in any given startup could be lowered.
The current approach to valuations includes basic tools such as formulas, calculators, and spreadsheets, all supported by a VC’s intuition. While these tools may never completely vanish as part of a VC’s due diligence, they lack the predictive capabilities and rich-picture approach a simulation can provide.
Simulations aren’t just theoretical, futuristic ideas. Growth Science, a data science firm founded by Thomas Thurston, already uses them. Thurston’s simulations, he claims, correctly predicted that Snapchat, Uber, and Airbnb would be big and that their accuracy when predicting whether or not a company will still exist in five years is 66 percent.
“Comparing our ‘quant’ [quantitative] approach to traditional VCs is like comparing a qualitative stock picker with an algorithmic hedge fund,” says Thurston. “Our process is mechanical. We use data, math and rules to try and isolate the specific percentage probabilities that any business will hit our goals as investors. We invest based on quantified probabilities, rather than intuition.”
The implications for startups are huge.
According to Thurston, Growth Science can “not only make probability-based predictions about whether a startup will succeed or not, but they also let us run ‘what if’ scenarios and to test-drive multiple strategies. This way, we not only know what’s likely given the startup’s current assumptions, but we can also identify trouble spots and help the startup course-correct to be more successful.”
Growth Science claims to draw on private and public databases across industries, and has “guided billions of dollars in organic growth, acquisitions and corporate venture capital.” according to the company’s website.
What further implications might the prevalence of startup simulations have?
Startup simulations have a broad economic impact if they are robust, realistic, and accurate. As the use of simulations becomes more widespread, the concept has the potential to alter how businesses are established, and how they grow beyond the early stages.
Startup simulations might not just predict future outcomes — they’ll create them.
Some details for this article were provided through a research project done as part of OCAD University’s Masters of Design, Strategic Foresight and Innovation program.