When talking about early stage company valuations, all opening paragraphs should start out with: it’s really hard. That’s why convertible notes are popular, why Y Combinator created the SAFE, (Simple Agreement for Future Equity) and why Draper is creating TATS (Tradable Automatic Term Sheets) effectively letting investors and entrepreneurs get around trying to value things today, and agree to try to figure it out later.

 

Normal valuation tools are tough (DCF, WAAC, FCF and all of the other acronyms that come along with valuing things), mostly because they rely on some proxy for cash flow, of which, early on, may be in short supply. Investors in early stage companies have to turn to other tools. If forced to rely on traditional approaches to valuation, we’re going to be cuffed with artificially low outcomes.

 

Often we think of valuation as the sum of the present values of an expected stream of cash flows. But since that can be tough for emerging companies, we can use a different framework to think about what else might be going on. To illustrate the idea, we can create a “Software Index” of publicly traded companies[1] and plot it against different variables & use Total Enterprise Value (TEV) / Revenue as our proxy for valuation.

First let’s plot valuation against EBITDA growth. There is little correlation. In other words, the market will tell a company with 0% EBITDA growth that it is worth 4.0x revenue, and will also tell a company with 300% growth that it, too, is worth 4.0x. Let’s try something else.

Well how about EBITDA margin as a proxy for cash flow[2]? Intuition would suggest the more robust a firm’s margins, the stronger the valuation. But that’s not what’s going on here. When looking at that same data set against EBITDA margin, the correlation isn’t just weak, it’s inverse! The stronger the margin, the lower the revenue multiple.

Revenue growth seems to be a more reliable driver of enterprise value in software companies. Here, we see a 37% correlation between revenue growth and valuation, which, frustratingly, has nothing to do with cash flow and seems puzzlingly decoupled from fundamentals. Naturally, the question is: why?

Here’s what may be happening. Venture Capital and Growth Equity investors look at cash generation as evidence that companies have run out of really cool ideas. If they had ideas, they’d be begging for capital to hire developers & sales teams and generating huge cash burn (or losses). If not, they had better turn into dividend companies, which give up their sky-high valuations. The best proxy for what will drive a software or early stage company’s success is revenue growth (or by extension, customer growth, or user adoption). If it’s a slick product, addressing a big market with a tough need, people will buy it. Once you have a critical mass of customers, you can focus on optimizing your operating expenses and start to manufacture cash flows. That brings us to a few other drivers of how investors might value an emerging company.

 

Revenue composition

A big addressable market is a foregone conclusion. Beyond that, most early stage companies are attacking trendy corners of the market: ecommerce2.0, marketplaces, enterprise software, security, FinTech etc. Most participants in these industries fall somewhere across a spectrum of ‘productization’ – to what extent the product is ‘fully baked’ or does it require considerable consulting & customization before it can be implemented and scaled. Fully baked solutions can fly off the shelf with low marginal costs, while the latter is takes time and expense to replicate for each and every customer. That brings us to …

 

Gross Margins (GM)

For emerging tech companies with recurring style revenue streams, this is probably north of 80%. Capital deployment needs to be efficient and easily scaled. With a few million dollars of capital a software company can scale a 90% GM idea coast to coast. It’s hard (and capital intensive) to pump out 10x growth on 35% GM. This is key to driving a good…

 

LTV / CAC Ratio

Early stage companies with good ideas tend to hemorrhage cash. Investors will use other tools to get a handle on the underlying strength of the ability to scale an idea. David Skok at Matrix Partners, a VC firm in Menlo Park, does a great job walking through the concept[3]. The idea is to understand what each marginal customer is worth over its lifetime compared to the marginal cost to acquire the customer. The stronger that ratio, the more efficiently the capital investment can be scaled (and more attractive the opportunity becomes to investors). If a software company wants to use Workday to benchmark its TEV at 9x Revenue, it had better have a CAC ratio north of 6x. Generally, LTV/CAC ratios should be north of 2.50 - 3.0x. A great article from Blossom Street Ventures[4] makes the case that 8.58x takes you public. This helps us think though…

 

Fund Economics

Early stage investors like to target 4-5x returns on their investment. IRR is one way to think of it, but that’s gets fuzzy the longer an investment is held. So plugging a top-down forecast into a return model that kicks out 4x return is a nice way to start backing into the framework of an addressable market that makes opportunities attractive to investors.

 

None of these have all that much to with cash flow on the front end – which can make ‘price discovery’ a challenge. And relying on one of them alone – a monster forecast, or a $30B addressable market – is probably not enough to convince a professional investor that a company should trade at 9x. But taken together: strong ‘productized’ recurring revenue composition, benchmarked gross margins, high LTV / CAC ratios, and a visible roadmap to generating 4.0x exits, we can build a case for market-driven valuations against public & private comps with similar characteristics. Ultimately, revenue growth is the most powerful predictor of early stage company valuation.

 

[1] CRM: MSFT, ORCL, CRM, SAP; Security: CA, FEYE, FTNT, IMPV, PRGS, RHT, SYMC; HR/HCM: CSOD, WDAY; Other: BSFT,CVLT,PTC,SPLK

[2] Not much changes if we look at FCF instead: correlation goes to 0.014