Posts

Even in infrastructure, many of the big data technologies driving increases in IT spending like Hadoop and Cassandra were developed by consumer internet companies (Google, Yahoo, Facebook, Amazon, etc date: 2012-10-17

The first mobile phones were purchased by corporations and given to employees. Thirty years ago, most people used computers at work but not at home. Most of the innovation flowed from the enterprise into the home. Today, it’s very much the opposite.

The big trends in enterprise trace the opposite movement both at the software layer and the device layer: consumerization of IT means using consumer channels to acquire customers and bring your own device (BYOD) means 66% of employees bring their own devices to work.

You’re walking down the hallway at work from one meeting to the next. A colleague or report stops you en route, asks for a minute and presents an important problem. It’s easy to respond with “let me think about it” and duck into the meeting. In that half-second, all the responsibility of the decision has been transferred. Unlike a minute ago, you have the monkey on your back.

The challenge with these situations is two-fold. First, you’re unlikely to have enough information to make a decision. Second, no time has been allocated to solving the problem.

One way of measuring the efficiency of a company’s revenue model is to benchmark revenue per employee. Google and Facebook, the two most efficient companies, generate $1M per revenue per employee per year.

Setting aside those exceptional cases and focusing instead on SaaS companies, the typical average revenue per employee is about $190k to $210k per year. The histogram above shows the ranges for publicly traded SaaS companies.

Financial statements are a Rosetta Stone for startups. They reveal the strategies and the tactics of how to bring a product to market. These are the ten metrics I look at when sifting through a startup’s operational model, whether when considering an investment or in a board meeting.

1. Revenue growth indicates how quickly a company can grow under the current way of doing business. The top line shows whether the market affords steady growth (SaaS) or lumpy revenue growth created by the long sales cycles of big customers (Telecom) and whether the company must sell one product or a collection of complementary products. The revenue growth projections indicate the potential of the business.

On January 8, 1966, the New Yorker profiled Buckminster Fuller for the first time. During a trip to a Maine island with the journalist Calvin Tomkins, Fuller said something tremendously prescient:

Fuller proposed a worldwide technological revolution…[that] would take place quite independently of politics or ideology; it would be carried out by what he calls “comprehensive designers” who would coordinate resources and technology on a world scale for the benefit of all mankind, and would constantly anticipate future needs while they found ever-better ways of providing more and more from less and less.

The traditional theoretical price demand curve is often drawn like this. The chart makes two points: there is some relationship between price and demand / revenue opportunity, and customer segments underpin that relationship. Each segment demands different products to satisfy different needs and presents a different revenue/profit opportunity.

Even if the details are very hazy, price demand curves are useful tools to inform product strategy and prioritization. To make PD curves useful requires marketing research. It means identifying customer segments, estimating the the demand of each segment, quantifying the investment required to build the right product, and uncovering the costs of acquiring and servicing customers within that segment. After all that, there’s also a process of testing various price points to get a better sense of price elasticity. All of these are useful in the process of establishing product market fit.

Raising money for a startup is expensive. The typical legal fees for a Series A are about 1% of the total money raised: roughly $40k on $4M. Of course, this doesn’t factor in the time for the process and the dilution of the investment.

But if your startup is considering an IPO be prepared to pay eight times as much in fees. Across 360 venture backed technology IPOs in the last 10+ years which on average raised $107M, 8.8% of the dollars the startup raises in the initial public offering is paid to investment banks, accountants and attorneys.

When startups are acquired, there are many considerations in accepting an offer. Does the vision of the acquirer fit the startup? Will the startup operate independently or be integrated? What is the price and structure of the transaction?

Most of these questions have to be answered through extensive conversations with suitors. As for the structure of the acquisition, there’s data that can be used for benchmarking. I’ve assembled about 2400 M&A events of venture-backed technology companies since 2000 to compare the fraction of the total consideration which is stock and cash.

Financial statements are a Rosetta Stone for startups. They reveal the strategies and the tactics of how to bring a product to market. These are the ten metrics I look at when sifting through a startup’s operational model, whether when considering an investment or in a board meeting.

1. Revenue growth indicates how quickly a company can grow under the current way of doing business. The top line shows whether the market affords steady growth (SaaS) or lumpy revenue growth created by the long sales cycles of big customers (Telecom) and whether the company must sell one product or a collection of complementary products. The revenue growth projections indicate the potential of the business.
2. The Net Income, aka bottom line or burn rate if negative, is the revenue minus all the costs incurred. Net Income dictates the minimum amount a startup needs to raise to become profitable. By comparing Cash, Net Income and Revenue, I can calculate when a startup will need to raise its next round, what its financial profile might be when it does go to market and get a sense of follow-on financing risk.
3. Gross margin is a measure how expensive it is make the product. It’s calculated by taking the revenue and subtracting all the COGS (costs of goods sold), which in software businesses includes serving and hosting costs, software licenses used, and revenue share agreements in the case of ad networks. Most software businesses have gross margins of 80% or more, which make them very attractive. On the other hand, grocery stores operate at 1–3% gross margin. Gross margin is the glass ceiling of profitability because the net margin can never exceed the gross margin.
4. Contribution margin measures profit per unit, without considering fixed costs. To calculate contribution, take the total revenue generated by selling one unit and subtract the variable costs to sell that unit. Selling and marketing costs tend to form the bulk of contribution margin costs in software companies. The greater the contribution margin, the more profitable the business on a unit basis, the more sales and marketing dollars can be spent to acquire customers and fuel growth. Contribution margins range from 5% to 25% depending on the sector. Of course, contribution margin has to be put into context. The contribution margin of producing a kilowatt of energy from a fusion reactor might be nearly 99%, but the fixed costs are measured in the billions. Contrast that with a 15% contribution margin ecommerce business which requires only $25M to become profitable. Which is more attractive?
5. Customer acquisition payback period or sales efficiency is a gauge for how aggressive a company can be marketing and selling its services. The longer the payback period, the greater the risk that a customer churns and the marketing dollars paid to acquire the customer are lost, and vice versa. The most efficient businesses I’ve seen recover their customer acquisition cost in 6 months. A 12 month recovery window is more typical. In SaaS startups, the payback period tends to be the driver for moving from monthly pricing to annualized contracts because it eliminates the profitability risk of a customer.
6. Churn quantifies the revenue potential of each customer. A 3% monthly churn rate, which is typical of public SaaS companies, implies most customers will only hang around for 2 years. The greater the churn, the more challenging revenue growth becomes over time. This often means a company will stimulate demand using paid acquisition, decreasing contribution margin and impacting profitability.
7. The single biggest expense for most startups is salary. By looking at salaries across functional areas, I can get a sense for how a startup pays its employees relative to market rates. Low salaries could spell employee retention questions in the future. Excessive salaries reduce the company’s runway. I compare salaries to a set of benchmarks across venture backed companies as a litmus test.
8. Sales quotas provide indications of how easily the product is sold and how well run the sales team is. Inside sales quotas vary from $150k to $600k per year and enterprise sales quotas vary from $1M to $2M, depending on the product. At the early stages of a startup, I tend to value consistency: smaller deal sizes but more predictable deal velocity (number of customers closed per week).
9. Non-personel marketing spend is the most significant controllable expense in a business. It typically includes ad spending and event spending. This expense bucket can be turned on and off from month to month unlike salaries or rent. It’s important to me because it provides an indicator of how well understood the marketing process has become. For most startups I’ve seen, demand generation budgets range from 5 to 20% of total expenses. The optimal ratio depends on the payback period.
10. Revenue per employee. The beauty of software businesses is their leverage. Google’s market cap is 40% larger than Walmart but it has only 2% the size of Walmart’s employee count. Revenue per employee is a measure of how efficient a business is in using technology to bring their product to market. Some sectors and products intrinsically need more people to be sold.

Bitcoin has captivated the imaginations of many with its quasi-anonymous, hyper cost-efficient payment network. The potential for Bitcoin to change foreign exchange is hard to overstate. In the same vein, the technologies that enable the internet of things (IoT) like Bluetooth Low Energy and Apple’s Beacons and Electric Imp’s infrastructure will transform the way we interact with the physical world to something akin to the mall in Minority Report.

The startups that bring Bitcoin and IoT to the mass-market won’t declare themselves Bitcoin companies or connected devices companies. I’ve never seen a computer marketed as a TCP/IP machine or a bank that promotes the ACH compatibility of its checking accounts or a social network trumpet its H.264 codec support. I doubt I ever will because supporting a protocol isn’t a value proposition that attracts end users.