Top metrics for evaluating startup health

What a Star Wall Street Analyst Learned About Evaluating Companies

After a decade on Wall Street and two more as a venture capitalist, Mike Kwatinetz, of Azure Capital Partners, has developed an approach to evaluating businesses that has helped him spot both promising start-ups and breakout companies. Rather than get bogged down in trying to understand hundreds of data points, he asks entrepreneurs to create a single spreadsheet showing little more than a handful of key performance indicators (KPIs). Critical to the exercise is going back two years on all those metrics, and projecting them one year into the future.

“Too often, founders want to show me 100-line spreadsheets, with line items for predicted monthly expenses plotted out for the next three years,” says Kwatinetz. “It’s not important to me what your phone bills will be in [three years in the future]. You want to paint a picture that shows whether you’ve got a good business, even if it’s not profitable now.”

A spreadsheet to validate your basic business model

Something like the view below gives Kwatinetz a quick sense not only of where the company is, but also whether the business model is progressing toward its ultimate profitability goals — without wasting a lot of time trying to predict things that are neither consequential or controllable enough to be worth the effort. Even if you could know for certain what employee dental benefits would cost in three years — which you can’t — it wouldn’t be a big enough number to matter. Including it would merely take time away from what you’re after: perspective.

Download a working version of the KPI spreadsheet.

What exactly belongs in this spreadsheet may vary slightly depending on your business. A hardware company, for example, would need a line to reflect inventory management, whereas a company that sold entirely via self-service e-commerce would not have to reflect sales-related expenses. But in all cases, Kwatinetz insists on the following five figures: revenue, cost of goods sold, gross margins, marketing expenses and operating profit.

Cutting out the noise

This particular set of metrics works because seasoned experts can quickly spot the underlying patterns that matter. As an analyst in the 1990s, Kwatinetz honed this approach as he made early calls predicting the rise of Microsoft, Dell and others. He used it as a VC to identify promising investments like VMware and Bill Me Later (acquired by eBay).

“It’s a helpful exercise because it shows you a story line, and it always leads to a great discussion,” says Eric Grosse, co-founder and CFO of Chairish, a home decorating marketplace. “It reminds me of the famous Mark Twain quote: ‘I didn’t have time to write a short letter, so I wrote a long one instead.” (Azure provided seed financing for Chairish, and invested in a subsequent round as well.)

What’s your contribution margin?

One of the most important insights — maybe the most important — that results from this analysis is understanding your underlying contribution margin, says Kwatinetz. “Everyone understands that most start-ups will lose money at the start, in the belief that when they scale they will become profitable. Well, whether that happens is a function of your contribution margin,” he says.

While definitions of contribution margin can vary, for the sake of this exercise, Kwatinetz simplifies it to gross margin dollars (revenue minus the cost of sales, such as parts and factory labour) minus sales and marketing costs.

Since sales and marketing tends to vary with the number of products sold by a technology company, the contribution margin tells you the degree to which your products are helping you cover the fixed costs that are rigid facts of life — things like executive salaries, property taxes and general administrative costs for functions such as payroll and IT. If your contribution margin dollars never exceed these basic costs, your company will not have an operating profit.

Here’s an example. A fast-growing Series-B start-up has £10 million in gross margin profit, but spends £20 million on sales and marketing. That’s a negative contribution margin of 50%. That’s OK, because it’s not easy to win customers for the first time—and because hopefully those customers will buy from the start-up for many years without nearly as much marketing and sales investment.

Now let’s say it’s three years later, and gross margins have doubled to £40 million while sales and marketing have grown to just £30 million. That’s a contribution margin of 25%, enough to help cover those other costs. If gross margin grows to £40 million but so does the cost of sales and marketing, the company’s sales are still not “contributing” to reaching operating profitability.

Contribution margin targets

There’s no single right contribution margin for every company. Labour-intensive companies with high fixed costs will tend to have lower contribution margins, as will companies which sell low-margin products. Even some high-growth firms may have negative contribution margins. But a healthy Software as a Service (SaaS) company should have a contribution margin of 25% or more.

Historical data is the best antidote against magical thinking

Kwatinetz’ quick-look exercise differs from the norm in a few other ways, as well. For starters, most entrepreneurs focus on how these key metrics are trending in the future. Kwatinetz insists on at least two years of historical data. For starters, it’s the only way to factor in seasonality, which is especially important for any consumer-focused company, particularly e-commerce plays.

Maybe more important, showing actual historical data will quickly expose any “magical thinking” by management. “I call it the ‘and then the miracle happens’ phenomenon, where lines start moving up and to the right based on some unproven assumption about what the future holds,” he says.

He also advocates for quarterly rather than monthly data for these metrics. Not only does this smooth out lumpiness that can obscure the long-term trends, it also makes it possible to display three full years of data legibly on a single slide.