Can An Algorithm Help Start-Ups Get Credit?

The team at CB Insights is great at delivering high level data on trends in the private funding market. I use their data frequently in my analyses.

On Tuesday they announced a private company scoring product called Mosaic.  The product seems designed as a proxy for the Dun & Bradstreet credit reports that business partners use to evaluate established companies as counterparties.  The CB team believes the reports will be used to improve a start-up company’s outcomes when trying to attract vendors, lenders and investors.  If it works, start-up companies will get better access to growth resources.  This is an interesting idea and it will be fun to see how it evolves.

The CB team built Mosaic algorithms from digital “signals” such as number of job postings, social media traffic and sentiment, product announcements and web traffic.  The Mosaic score will be broken down into Momentum (social traffic), Market (sector trends) and Money (how much raised and from whom).

Screen Shot 2015-08-27 at 1.48.47 PM

The most valuable of the three score components over time may be the “Market” score.

The Momentum score will favor companies that have great PR, not necessarily great product.  If vendors / lenders begin to use that score, companies can work to game it with increased PR effort.

The Money score is hard to do well without private P&L data that CB will not possess.  Vendors & lenders already know who a start-up’s VCs are, how much they raised and when.  I am not sure the Money score will be impactful for lenders as a bank can get actual company financials from the company applying for credit.

The Market score, however, can quantify what lenders, follow-on investors and others do instinctively from being in the idea flow.  They make a decision about the attractiveness of the sector.  The flaw with instinct is we humans overweight the last data point, the recent flame out or last week’s successful IPO.  If the Market score can deliver a medium-term metric targeted narrowly enough to be applicable this can lead to better lending / investing decisions and more access to capital for successful companies.

I think Mosaic is a great effort by the CB team even if it hits only part of its intended mark.  My guess is that among the targeted audiences of vendors, lenders and investors we will see different usage.  If the scores prove valid over time, executing vendor agreements might end up being a lot easier and faster for start-ups.  Lenders will still rely on company P&L data, but might get better at sub-sector allocation within tech. Follow-on investors will start using Mosaic today. Any reliable screen that can help investors speed the process of getting in front of likely winners will get traction.  We will know it really works when someone markets a fund that buys private company stock based on Mosaic algorithms.

[Contributed by Mike Mackeen, originally published on Raising the Next Round]

Posted on August 27, 2015 in Capital Markets, Financing Trends, Insights