The following post was written by Bart Bohn, Director of ATI-IT/Wireless
Recently ATI brought on several new interns for the summer – most of them are graduate students from the University of Texas at Austin. Since they are so heavily involved in due diligence of companies applying to join ATI, we spent part of the orientation session on performing due diligence. One topic generated a good conversation and I want to share it – financials. I told them that whenever I see the hockey stick financials slide that I just drop it into the shredder. The output of the financial model doesn’t matter; it is all about the assumptions that go into it. In every good financial model, there are really 3 – 5 key assumptions that drive everything – typical ones include:
- Customer acquisition costs
- Revenue per customer
- Cost per unit (sometimes called BOM – bill of materials)
These three make up the “unit economics analysis” that every company must understand. Until each unit sold produces a profit, there is no point in moving forward with other analyses. The other typical ones include:
- Sales calls and conversion rates per sales person
- Revenue split with partners or distribution fees with distributors or retailers
- Revenue / staff (more for service oriented companies than product ones)
- Cost per product generation (think designing, taping out and manufacturing chips or massive software releases)
- Customer growth rates – can break down into either count of customers or spend per customer
Each business is different, but each one will have only a handful of metrics that matter – and they frequently blend operational and financial issues. This post is biased towards B2B, but either these or similar ones apply to B2C as well. One of the hardest parts about generating decent assumptions is having defined your market correctly. Many entrepreneurs identify a very big market because it is measured in billions of dollars, and they think investors need markets that large. While that is true, what is more valuable is segmenting your market to the point where the above assumptions are generally consistent from potential customer to potential customer. This does require the entrepreneur to have extensive knowledge about their target market, almost to the point that they must have been part of it in the past – or they could fake it really well at a trade show.
My personal bias in evaluating very early stage companies is to generally relegate financial analyses to second tier – aside from unit economics, as I am a qualitative decision maker. This corresponds, I think, with being an Intuitive on the Myers Briggs framework. Although I have undergraduate degrees in chemical engineering and finance (with a minor in math) and a MBA focused on finance, I make decisions based on qualitative, narrative oriented methods. The reason is that I have built so many models, especially ones with feedback loops, that I can create numbers to support almost any conclusion. I can frequently take an entrepreneur out behind the financial modeling woodshed, and have done so enough times — when in a grumpy mood (sorry) — that I have become extremely cynical. I suspect many other investors are the same.
The models that I do like have real world data points to justify each assumption – and how the assumption changes over time. One entrepreneur did a customer acquisition cost analysis and used cost of sales for Salesforce to justify the cost of sales curve as his company scaled. That is good – while his company may not perform like Salesforce, understanding the potential shape of the curve and resulting operational implications (hiring requirements, cash flow timing, etc.) is extremely valuable and shows the entrepreneur is really thinking through the business.