Predicting the Trajectory of Very Early Stage Companies

[Sorry this is a longish post, but I kind of felt the need to build to the question I wanted to raise]
It’s widely acknowledged that initial startup costs for many internet-enabled companies (SAAS, consumer web, mobile apps, etc all fall into this bucket) have dropped dramatically in the last decade. As a result, many companies can get off the ground and make significant progress on modest amounts of capital… often well under $1M initially.
Countless companies you’ve never heard of got started on a few hundred thousand dollars or less. Most never become anything, a small portion become small to medium sized companies, and a tiny handful become hugely successful companies. It’s easy to think of the ones that got big… like Facebook which started off as a product in a dorm room, and grew as a startup with a $500K investment from Peter Thiel in 2004, entered the realm of VC-backed companies in 2005, and of course today is a worldwide phenomenon and very successful company.
The chart below isn’t meant to be a highly precise analysis, but rather a graphical depiction of broad “envelopes” of outcome and capital investment. There are always rare exceptions, but the vast majority of companies that ultimately become $1B+ outcomes raise at least tens of millions of capital (if not significantly more) to get there. Similarly the vast majority of companies that only raise ~$1M or thereabouts in capital result in outcomes of substantially less than $100M.

The question of what makes a great seed stage investment has two answers, one absolute and one relative to the investor you’re asking. Outcomes in the upper part of any of the envelopes above could be considered absolute successes, but relatively speaking these may not be considered success depending on the type of seed investor. A good outcome (i.e. return on investment of 5-10x+) in the green box may be considered a win for an angel investor if the company remains capital efficient, but it wouldn’t be for most VC funds (small or large). A good outcome in the red box might be a win for a small VC fund but probably not for a large VC. A large VC obviously seeks outcomes in the top part of the blue box. These relative answers explain the motivations of different types of investors actively pursuing seed stage companies, as Chris Dixon and plenty of others have discussed.
So for many in the VC and startup ecosystem, none of this is news. What I think remains very much unresolved in this world of lower startup capital requirements is the question of how well investors and entrepreneurs can prospectively predict the potential scale of outcome. I’m not talking about predicting the likelihood of success (obviously important), but rather predicting how big the company might ultimately become and how much capital & time it might take to get there.
There are a range of factors that can help in trying to assess the probability of actually building a company of significant scale:
  • Size & Attractiveness of Market Opportunity – some are obviously big and some are obviously small, though many not entirely clear
  • Caliber of Team – never a precise science judging a team, but well understood ways to evaluating a group of co-founders based on their prior accomplishments, raw intellectual horsepower, charisma, etc. In addition to raising the chances of success, small or large, an exceptional team also usually increases the chance of creating something very big.
  • Founders’ Stated Ambitions – most entrepreneurs like to think big, but sometimes they make it clear to seed stage investors that their goals are more modest. Entrepreneurs also make a continuous evaluation over the life of a startup as it reaches various value creation points.
  • Scale of Competitors – if other startups in the space have already achieved real scale a new entrant might also be able to, but again hard for those startups truly pioneering a market or new class of product
Companies where one can prospectively see high probability of a large scale outcome are attractive seed investments for large VCs, provided they believe there’s also a good probability of success. It’s not uncommon for VC’s to write $250-500K “blank checks” to entrepreneurs they know working on a concept that fits this description. Similarly those that are probably not venture scale outcomes (i.e. the green box above is a best case scenario) may receive angel investment but typically not seed funding from VCs.
But what happens to everything in between those two extremes? What’s the shape of the probability distribution, i.e. do 10% of startups obviously have large scale potential and 10% obviously have small scale potential, with the remaining 80% hard to assess? Or is the distribution more even?
Take Twitter just as an example… it started life as a side project within a struggling startup called Odeo. In 2006, Ev Williams returned what was left of Odeo’s capital to its investors (in fact made them whole out of his own pocket) and was widely praised for doing so. He (and collaborators Biz Stone and Jack Dorsey) was rather uncertain about what sort of scale Twitter or other projects within Odeo’s successor corporation (Obvious Corp) might ultimately achieve, and freely admitted it at the time.
So what happens to the startups that have decent potential for success, but real uncertainty about scale of outcome? Do most of these get funded and launched or do most die on the vine without ever taking a shot?
Some large VCs deal with this segment with a “portfolio” approach (or less generously described “spray and pray”), by making a large number of passive seed investments and investing larger amounts only in the small handful that prove out A) some greater potential for scale than at inception and B) some continued probability of success. Some of these companies undoubtedly receive angel funding (e.g. Facebook
example above).
A very small portion of VCs have specifically crafted their investment strategy in part to cope with this uncertainty. For example, Josh Kopelman has described First Round Capital’s outlook on scale of outcome with the express vs local train analogy. FRC can seed a company and if it exits at $10M or $40M, that works fine for the firm given their strategy and fund size. If it has a chance to take in more capital, but shoot for a $200M+ outcome that’s fine too. His point is that this funding path looks like a “local” train whereby you can get to a faraway destination, but also have options to disembark at stops along the way. By contrast seed investments from large VCs can be likened to an “express” train… i.e. they can work just as well and maybe get you there faster, but only if all parties involved are committed to the faraway destination at the outset.
What do you think? How easy or hard is it to prospectively predict potential scale at the seed stage? For those that are difficult to clearly predict scale, what happens to them today? How ought investors (either individuals or VCs) approach these opportunities in the future?

Lee Hower

I’m an investor, entrepreneur, and helper of technology startups. I’m currently a General Partner of NextView Ventures, which focuses on seed stage internet-enabled businesses. I co-founded NextView in 2010 with my partners Rob Go and David Beisel. I started in the VC business as a Principal at Point Judith Capital, an early-stage firm. I joined PJC in 2005 and served as a Principal at the firm through early 2010. During this time I co-led investments in FanIQ, Sittercity, and Multiply and sourced investments in Music Nation and NABsys. Prior to becoming a VC, I was a startup guy myself. I was part of the founding team of LinkedIn, and served as Director of Corporate Development from the company’s inception through our early growth phases. Before that I was an early employee at PayPal, and worked in product management and corporate development roles through the company’s IPO in 2002 and subsequent sale to eBay later that year. I went to college at UPenn and received degrees from both the School of Engineering and Wharton School of Business.

    • Great post.

      When we started out with, it was called by some a "lifestyle business" because the market was small. We later realized that our business model (taking a commission on online bookings) was simply not possible given our limited capital (because you need to do a ton of direct sales to get Bed & Breakfast owners to use your software to track their room inventory).

      So we switched to developing Postling, a social media management tool for businesses to do all of their blogging, facebook, twitter, and flickr all in one place. It's a bigger opportunity and got even bigger as we thought about what Postling would look like if used by 10,000 or 100,000 users. We'd have a content network with millions of page views across millions of social media content, detailed metrics on everything, and metadata about the businesses and their audiences.

      My point is that with a strong team, you never really know what will happen as the team pivots and their ambitions change as new opportunities reveal themselves.

    • Lee

      Thanks David. I wanted to raise the question and get a sense of what others believed, so great to hear your example. My own hypothesis is that the segment of companies like yours where the scale of oppty can both change or otherwise be hard to predict initially may be rather significant now that initial capital requirements have dropped so much.

    • Great post Lee. Very thoughtful and thorough.

      I wonder how many companies go unfunded in the middle bucket, and how many other companies do get funded in that bucket, but do unnatural things to try to get to the third bucket.

      Definitely a gap there.

    • Lee,

      Great post, and I'll be borrowing that graphic with attribution, iydm.

      I'd like to suggest that "number and size of adjacent markets/opportunities" is a better predictor than caliber of team for the potential of a startup to be big.

      As you pointed out, likelihood of success is different, though important, than potential to be big. I think team caliber contributes mostly to that second factor.

      As most startups pivot at least once in their early stages, having adjacent market segments that are also big, or can be strung together in a way that isn't business model gymnastics, is key when that first group of customers doesn't work out quite like you thought they would.