by Beth McKeon, Fluent CEO

$621B was put to work in startups last year alone, up 149% from the previous year.

All that money yet no one in the industry has complete data from beginning to end on these companies, either the ones that fail or the ones that get funding.

If this were true in the medical field, we’d have no evidence to explain the reasons people lived or died. We wouldn’t know how to train doctors or equip hospitals or prescribe a course of treatment.

Missing the much needed causal evidence of what leads to success, the industry has adopted proxies for data-driven decision making.

The most common proxies are “who you know” and “what you look like” (factors unrelated to the business itself).

Entrepreneurs are optimizing for pitching skills and trying their very best to “look” like the kind of founder that gets funded. These efforts are completely unrelated to the skills needed to build a company. Not to mention some founders will never look like the entrepreneurs who currently pull down the best and biggest deals.

When investors pattern match off of the winners (survivor bias), it forever replicates the inequalities of the system.

We all know these biases drive the industry and have since this asset class developed. What hasn’t been talked about is how our current attempts to solve this problem have also failed.

First, the industry began to acknowledge that almost all of the capital was managed by white men. Who they knew and who they trusted clearly also favored white men.

The solution appeared to be as simple as seats at the table: If we want more women and people of color to access capital, we need to have more women and people of color managing capital and running funds.

This necessary change is wonderful … However, the stats haven’t changed. Why?

No matter who is sitting at the table, investment decisions still use the same flawed proxies and still perpetuate the same biases.

As Rukaiyah Adams at Meyer Memorial Trust says, once given an “invitation to the cookout”, even women and people of color must maintain the systems of power as they currently exist or they’ll get kicked out.

The structure itself requires that the biases perpetuate themselves, no matter who writes checks. The necessary solution to change who gets funded is to restructure the system to value data-driven decisions.

Let’s look at another industry where this has already played out, and where data won.

Recall what the car buying experience was like before Carfax and online comparison shopping. The buyer walked onto the lot and had to decide how much to trust the salesperson, both about the quality of the car and the value of the car against the price.

Furthermore, some people got access to relationship-driven (who you know and what you look like) discounts, which meant women and people of color were likely going to get a relatively worse deal

What solved this structural problem within the car industry? Data and transparency.

When the car buyer could access as much information about the quality and value of a car as the salesperson, the car buying experience normalized. The trust in the transaction was evidence-based. There was no need for conferences about diversity and inclusion, and no need to set up special dealerships for specific kinds of customers. Eliminating the information asymmetry was the needed structural change.

Data is the solution to the VC industry’s problems. Funding the best companies, based on data and not precedent, solves the equity problem.

Here at Fluent, we’ve already seen this work. In an analysis of how accelerators use the Fluency Score alongside their typical methods (application + interview) to select their cohorts, this data has consistently led to more diverse and equitable cohorts than would have been selected otherwise.

The Fluency Score is just the beginning, just the earliest signal that data tracking and reporting on startups from launch through exit has the capacity to transform our industry.

Why does all of this matter? Everyone benefits when the decision to fund a company is based on data. This professional asset class needs professional tools or else it will continue to walk an infinite loop of biased bets, instead of the best, data-driven bets.