“Last night’s game, like Saturday's, ended with a losing-team player disconsolate in the dirt, but this time without an attached ruling to talk about. Kolten Wong, a ninth-inning Cardinals pinch base runner, was cleanly picked off first base by the Boston closer Koji Uehara, for the last out of the game. No excuse: Sox win, 4–2, knotting the series at two games apiece. “
--Roger Angell, The New Yorker Blog, October 28, 2013
Scoring is the heart of the competitive game. Without it, Roger Angell’s meaning is gone.
Now, baseball isn’t scoring, of course, and no scorekeeper ever made it to the Hall of Fame. That’s reserved for the great players who inspire us with skill, courage, grit – American virtues that built this country.
But without the humble scorekeeper, and his offspring, the statisticians and analysts, it is hard not only to cite the deeds of the heroes, it’s also hard for those same heroes to know what to do.
It isn’t a game if there isn’t a score.
For entrepreneurs and their companies, it’s not a game without a score, either. While a baseball score certainly turns into cash in many ways, none of them are as direct as sales and sustainable cash flow in a venture.
And since the state of most startups is to be underfunded with more to accomplish than is humanly possible, the appearance of the unexamined hypothesis is almost guaranteed. Its cousin, the vanity metric almost always comes to the party, unwelcome and unnoticed, too. You may find them intertwined like moonstruck teens and just as hard to separate.
An unexamined hypothesis – ones that are used to plan with – are very dangerous. It can be many months or years before they’re proven to be wrong at a very high cost.
The vanity metric – “1.8MM visitors – 150,000 customers,” – are metrics that sound good – even great – but don’t tell us anything we can do anything about. How did we get them? Are they profitable? Or spending anything? They’re ok for PR, but offer no guidance.
Hypotheses are a central concept in a successful business model and accurate metrics are a requirement. But a successful hypotheses must be falsifiable – that is, it must be stated in a way that it can be factually disproven.
Real metrics need to reveal not only whether a hypothesis is false or true, but also how it can be actionable.
The customer acquisition component of the business model drives the entire business. Every entrepreneur wants to know how long it will take to and what it will cost to obtain customers, how long they stay, what they’ll spend, and over what period of time how much contribution margin they’ll contribute. (In some cases, a product spreads by work of mouth, “goes viral,” but less often than most entrepreneurs would like. And even in those cases, some acquired customers are better “carriers of the virus” than others. This tends to apply more to B2C companies than B2B.)
When the product is ready, most companies won’t have had enough sales experience to really know what their initial acquisition costs are. So, enter the hypothesis.
An entrepreneur might surmise, from the best data available to her, that their acquisition cost will be $300 per customer, and that there are many, many customers available to be acquired.
A customer, based on beta tested pricing, will deliver an annual contribution margin of $500, paid monthly. After about 8 months, that customer will generate positive cash flow of about $41 per month.
Spending $30,000 in month one will generate 100 customers, meaning about $4,100 in cash.
Spending $30,000 every month for a year adds $4,100 to the cash flow each month. After a year, the cash revenue is about $49,000 per month.
It’s tricky to calculate what an acquisition cost really is. What activities do you ascribe to it? Costs? Are you under-allocating your fixed costs to acquisition?
From the first day of revenue, a critical examination of acquisition costs – and the time acquisition takes using veracious cost based methods – is a necessity. This disciplined monitoring of ongoing acquisition costs allows for accurate course correction and will lead to diagnosing some important aspects of the business model.
Not all customers are equal. Some are described by where they came from (acquisition sources), others by their “type,” (demographic characteristics) others by their business (NAICS code) or their business size. (under 500 employees). There are many variations and combinations that make up segments.
This is important because in most situations, acquisition costs rise, and the best performing subgroups (segments) often are the smallest and least able to grow. In addition cohorts, (a group of customers acquired together at a fixed time in the past, i.e., “June, 2013) often change behavior one from another. This may be due to changes in the original promotions, offers, or product features.
This is just one example of the need to prove out a hypothesis and be wary of vanity metrics. There are many more areas to apply this to, including growth planning and control metrics around increasing production, customer service and fixed costs; examinations of unit economics to determine contribution margins; and planning for fixed versus variable expenses.
Tim Keane is an entrepreneur and investor who’s also taught entrepreneurship and consults with growing companies.