Success projection for startups

December 5, 2012

Startups often get so consumed by their day-to-day challenges that they do a poor job actually projecting their longer-term success or lack thereof. As a result, many toil without knowing whether they're truly living up to the expectations they've set for themselves or not.

Even though there is a lot of uncertainty surrounding early-stage companies, they'd do well to build working projection models that give them a sense of whether they're on a road to success or growing too slowly per business metrics that matter to them most.

All startup stakeholders (founders, employees, investors, etc) ultimately gauge their long-term success by the company's anticipated valuation at the future point of time when they decide to liquidate their equity holdings, perhaps several years or more after they begin working on it. Therefore, any projection of success should place this valuation as its end goal and work backwards from there to derive the more immediate factors that go into achieving that success and whether those factors are on the right course.

A company's valuation is derived (at least theoretically in an efficient market) by the total amount of profit it will accumulate forever into the future, discounted to a present value. So, the first derivation from valuation is long-term profit, and since this profit consists of the company's expected revenues and costs, those come next.

Costs can be derived by projecting headcount and other operational expenditures, perhaps by studying public information about other companies that have developed in the same way as you expect your company to do so, providing rough guidance as to how your particular and total costs may pan out over the years.

Revenue can be projected through comparables as well, especially if you anticipate deploying a business model similar to that of a public company (e.g. you could study companies that make most of their money from display ads if that's what you also intend to do). However, you will likely learn the most from them about what kinds of per-unit rates they get from various monetization schemes (such as subscriptions, e-commerce fees, and advertising CPMs), leaving it to you to determine what kind of product usage volume you anticipate mapping against such expected rates.

If you are, like many consumer startups, intending to generate revenue from advertising, then the model's key questions are consequently: 1) how many active users do you expect to attain by a given date in the future, and 2) how active will they be per day, week, month or year. And this is because you'll need to estimate the size of an active audience that'll be at your disposal for advertisers. You may come up with innovative ways to increase ad rates, but your future revenue will be mainly tied to how large or small that audience becomes.

With this being the case, you need to focus on how quickly you are accumulating active users and increasing their engagement, and this is where the model starts to get concrete for even the newest of startups. The number of active users for a product at any given point, now and into the future, is determined primarily by its user acquisition rate (how many people are signing up or otherwise first engaging with a product per a given unit of time), its activation rate (what percentage of those people reach a point of appreciation for the product), and its retention rate (what percentage of those who activate continue to use the product repeatedly).

Each of these factors (as well as others that correspond to non-advertising-based business models) is unique to a given product, and eventually you'll need to project them all if you want to complete the model. However, even if you're at a beta stage with only 50 testers, you can start projecting them one-by-one, making assumptions about the rest. You won't have a lot of statistical significance with such a small user base, and it's probable that your key metrics will change as you address a larger market. But it'll at least give you a baseline from which you can judge movements towards or away from your ultimate definition of success (i.e. how valuable you want the company to become and how quickly). And it'll keep you honest about whether you truly have enough data to establish knowledge about the business's momentum, and if you do, whether that data reinforce or contradict your more subjective intuitions about how well the startup is doing.