Welcome to Wing’s third annual study of the changing nature of early stage financing. This is based on our “V21” analysis, which includes only technology companies financed by one of a select group of the industry’s top venture firms. As in years past, we have updated the study with a new year’s worth of data. We have also made some important improvements to our methodology. The curated V21 data set now captures the details of 6,546 financings at 3,389 companies invested in by one of 21 elite venture firms across 10 years.
The trends observed in prior years have accelerated and expanded. The Seed phase covers more and more of early-stage company development. Meanwhile, Series A financings and the companies receiving them increasingly look like Series B’s. As we have said before, “Seed is the new A, and A is the new B”.
What’s new are the unavoidable questions about where this is going. We are clearly in a different world now. 2020 will look more like 2001 and 2009 than 2019. This is the year that finally brings to an end of the trends that have been so dominant for so long. We will speculate on the future in the latter section of this paper.
Here is a summary of the key findings:
- Companies are consuming more Seed capital. In 2019, the median company had raised a total of $3.3M prior to raising a Series A, up from $3.0M in 2018 and 5.5x more than the $600K of 2010.
- The size of Seed financings has grown considerably. The size of Seed financings has nearly doubled in the last 4 years. The median Seed financing in the V21 sample was $3.9M in 2019, up from $3.1M in 2018 and 3.9x more than the $1.0M of 2010.
- Median Seed pre-money valuation is now in double digits. The median pre-money valuation for Seed financings in 2019 was $11.0M, up 17% versus $9.4M in 2018.
- Pre-revenue Seed financings are becoming scarce. Two-thirds of companies closing a Seed financing in 2019 were generating revenue, up from only 9% in 2010.
Series A Financings
- The size of Series A continues to grow, but more moderately. In 2019, the median Series A was $12.4M, up from 11.9M in 2018. While this is 2.5x bigger than the $5.1M of 2010, it represents a slower rate of growth than any year since a dip was recorded in 2016.
- This is also true of valuations. Median Series A pre-money valuation was $30.0M, up a modest 5% versus $28.5M in 2018.
- A distinct “jumbo” Series A cohort has emerged. In 2019, 43 companies closed a Series A financing greater than $20M, notably up from 28 in 2017. There were only 8 such “jumbo A’s” in 2010. This count specifically excludes Growth Equity financings which are often labeled Series A – our methodology identifies these financings and excludes them from the study.
- Pre-revenue Series A’s continue to be rare. 77% of companies completing their Series A in 2019 were generating revenue, up 10x versus 7.7% in 2010 but essentially flat since 75% in 2016. Perhaps we have approached an asymptotic limit?
Time Machine Analysis
Its tempting (and fun) to search for “look-alikes” between current-era financings and predecessor deals from prior years. Here is our best shot at it:
A 2019 Seed looks like a 2011 Series A
- Valuation (Pre-Money): 11.0M (Seed) vs 11.4M (Series A)
- Size: 3.9M vs 5.5M
- Round Number (Sequential): 2.1 vs 1.7
A 2019 Series A looks even more like a 2010 Series B
- Valuation: 30.0M vs 32.2M
- Size: 12.4M vs 11.1M
- Round Number: 2.8 vs 2.7
As we have seen in prior studies, “Seed is the new A”, and “A is the new B”. But for how long?
The Data: Updating the V21
Each year we update the V21 dataset and make improvements to the analysis. This year we naturally added a new year’s worth of data, which impacts some of the prior years as the longitudinal financing histories of newly included companies are added to the sample. We also made important methodological changes, most notably: shifting from “average” to “median” figures; removing obvious growth equity deals that happen to be labeled Series A or B; and removing “multiple” Series A’s and B’s, including only the first and / or most significant financing of the set. Much of this work had to be done by hand, leaning on the judgment of experienced investors, which is one of the things that makes the V21 analysis unique. For consistency’s sake, we reran the entire analysis all the way back to 2010.
For readers unfamiliar with the V21, here is a quick recap of the motivation and methodology. Others may want to skip ahead to the next section, “The Findings”.
Motivation for the V21 Analysis
The future success of an early stage venture is not yet known. To focus on the highest quality companies, you need a proxy for company quality that is observable in the present. The one we chose in creating the V21 is the quality of the investor.
There is a lot of research support for this choice of proxy. Analyses from academics, limited partners and even venture capitalists show that fund returns and top-performing companies are disproportionately concentrated in the portfolios of the very best venture firms. Of course it is true that even the best firms make bad investments, and mediocre firms sometimes make good investments. But it is also true that the highest quality firms vastly outperform the rest, with strong persistence effects. By focusing on the investment activity of the top firms only, we are able to create a sample that better represents the companies that matter.
The V21 starts with the identification of a group of 21 leading venture firms with a focus on the US technology sector. This is a subjective selection, performed by the partners of Wing, based on our deep experience in the industry. The firms in the sample are all best described as “Series A” or “multi-stage” firms. The group does not include Seed funds, although many of the firms in the sample make Seed investments as part of a broader strategy. This should not be interpreted as a negative statement about the importance of Seed funds, but rather the recognition that the portfolios of the leading major venture firms contain the highest fidelity signal. The best portfolio companies of the Seed funds are usually also captured in this sample, but with less accompanying noise. While we do not reveal the names of the 21 firms, it is fair to say that if we did, there would likely be no disagreement concerning 10 of them, broad consensus on another 5, and lively debate concerning the remaining 6.
Having selected our 21 firms, we next identified every company in which one of them has made a new Seed, Series A or Series B investment from 2010 to 2019. These “V21 companies” form the backbone of the data set. It is worth noting that a follow-on investment from a V21 firm is not enough to include that company in the set – the investment must be the first time that firm has invested in that company. We do not include companies where the financing was an obvious growth equity investment, even it was labeled “Series A” or “Series B”.
Finally, we collected information on all Seed, Series A and Series B financings for the V21 companies – their longitudinal financing histories. Some companies execute multiple Series A’s and B’s (e.g. Series A-1, A-2, etc). We have even seen an A-8! These additional “A-n” or “B-n” series’ might be associated with warrants, convertible notes, foreign investors, tack-on’s as well as true later-stage financings (sometimes many years later), among other things. In those cases, we identified which of these could most accurately be called the “true” Series A or B (based on timing and size) and did not include the others in our dataset. This is an inexact science but was done by knowledgeable investors using their best judgment. All in, the V21 data set now captures the details of 6,546 financings at 3,389 companies invested in by one of 21 venture firms across 10 years.
The focus of the V21 analysis is technology venture investing in the US. We specifically exclude companies in obviously divergent sectors and geographies. Some exclusions: no later stage investments, no growth equity; no life sciences; no China, no India. V21 companies in more closely aligned geographies and sectors, like Europe, Israel and Health Care IT, were left in the sample, but since the V21 firms are all US-based and technology focused, this is a relatively small proportion of the overall data set.
1. Size of Financings
The size of early stage financings continued to grow last year, especially for Seed financings. The median Seed financing in 2019 was $3.9M, up 26% on the year and 95% since 2015; the median Series A was $12.4M, up a far more modest 4%. Meanwhile, the median Series B grew 13% to 28.3M. The more rapid growth of the Seed financing has led to “multiple compression” versus the Series A over last 4 years (from 5x in 2015 down to 3.2x in 2019).
The “cross-class” comparisons are also instructive. The median 2019 Series A has now surpassed the 2010 Series B; and the median Seed has reached 78% of the 2010 Series A.
This year we started tracking the number of “Jumbo Series A’s”, which we define as greater than $20M. This used to be a rare occurrence and generally indicated something else was afoot – for example a growth equity financing in a mature bootstrapped company, or a corporate spinout. Because the V21 analysis is about early-stage financing dynamics, we have specifically identified such deals and removed them from our dataset. Even after doing so, we still see a distinct cohort of $20M+ Series A’s – 43 of them in 2019. This phenomenon started to come into its own in 2014 (when there were 33 such deals). It will be interesting to track the success rate of these super-capitalized companies, and to monitor the frequency of such transactions as we enter a less frothy phase of the market cycle.
Valuations continued to rise across the board in 2019, with somewhat sharper increases in the Seed and Series B. Median Seed pre-money valuations crested 10M for the first time, Series A’s reached 30M, and Series B’s reached 90M. It is tempting to assert a “Rule of 3x” (pre-money to pre-money between rounds) and indeed, this has pretty much been the norm for the last several years.
3. Sequential Number of Rounds
“Sequential Number of Round” in the V21 analysis indicates the chronological sequence of a financing. Is it the first? Then it is “round number 1”, and so on. For this metric we are still reporting the average: the step-function nature of this whole-number metric makes the median less informative.
The average “round numbers” appear to have leveled off for each round: Seeds are now round number 2, A’s are round number 3, B’s are round number 4. Whereas we used to think of the Series A as the “startup round”, with “2 founders and some slides”, it is now most typically a company’s third financing. The increase in the average sequential number for Seeds speaks to the phenomenon of multiple Seed rounds (“pre-seed”, “seed-plus”, “seed-2”, “post-seed”) as well as the massive number of “accelerator rounds”.
4. Cumulative Capital Raised “Prior-to”
Cumulative capital raised “Prior-to” is the total capital a company has raised prior to a particular financing. This is another metric where we report the average, as the median is distorted by a large number of “zeros”. This figure has increased massively over the course of the decade. V21 companies raising Series A’s in 2010 had raised an average of $600K beforehand. By 2019 this number was $3.3M, a 5.5x increase. With the size of Seed financings growing rapidly, it is reasonable to expect this figure to increase further in 2020.
Capital raised prior to the Seed increased 50% in 2019, to $1.2M. Thanks to the increasing prevalence of multiple Seed rounds, we expect this figure to continue to rise.
Capital raised prior to the Series B also increased 11.5% in 2019, to a whopping $20.3M. With the size of Seed and A financings on the rise, this figure should also be expected to continue to grow.
5. Years Since Founding
“Years Since Founding” continued to increase for all stages in 2019, most notably for the Seeds, where the median age of a company receiving Seed financing increased 41% to 1.72 years.Given that large increase, we should expect to see increases in the median ages of companies receiving Series A and B financing in future years.
6. Revenue Generation
With companies taking more time and raising more capital prior to every financing stage, it should not be surprising to see a rise in the proportion of companies that are revenue-generating. 77% of new Series A companies are now generating revenue, a modest 1-point increase over 2018. We may have reached some sort of asymptotic limit as this figure is essentially unchanged since 2017. The Series B cohort has also leveled off in the low 90%’s (92% in 2019, which is actually down a couple points versus 2018).
It is more surprising that a stunning 67% of companies completing Seeds are now generating revenue -- a 10-point jump over 2018. We can’t comment on the quality of that revenue, which undoubtedly includes many friends, relatives, former colleagues and accelerator batchmates as sources. But it is clear that many companies can now bring a product to market very efficiently and make substantial company-building progress during their Seed phase.
Venture Capital Market Cycle 101: Welcome to Freefall
The V21 analysis is by definition a retrospective. Most of us are more interested in where things are headed. It is important to remember that venture capital is a cyclical business, characterized by periodic booms and busts. In every up-cycle we hear claims that “things are different this time”, and in many ways things are indeed different every time. But the thing that is never different is that a down-cycle lies ahead.
We are coming off a sustained decade-plus boom the likes of which has never been seen. Most of today’s “venture capitalists” have never actually navigated a down-cycle, much less a total market collapse. The rules of the game are about to change.
I had the privilege of working alongside the great Arthur Patterson, one of the founders of Accel (my former firm), for nearly 15 years. After the dotcom / 9-11 crash, Arthur shared with us his four-phase framework for understanding the venture capital market cycle. This framework has since been adapted and expanded upon by several others, including Brooks Zug of HarbourVest. The last several weeks have seen us abruptly transition from the “Acceleration” phase of the cycle into the dreaded “Freefall”. This phase-change is always the most whiplash-inducing of the four. While every transition is unique, we have been here before – several times in fact.
What lies ahead in the Freefall phase won’t be much fun. The venture capital market will freeze as investors wait for the bottom and triage their portfolios. The pace, size and valuation of financings will all contract. Companies with high burn rates will need to reset them quickly if they expect to survive. Tourists (both investors and entrepreneurs) will go home. The grittiest entrepreneurs and venture capitalists will remain, they will depend on each other, and the work will be hard for both.
The 2021 V21 analysis will reveal a sharp reversal of some of the trends we’ve been documenting to date. Most notably, deal sizes and valuations will undoubtedly decline, as will the number of deals closed. However, company maturity at time of financing (age, percentage generating revenue) may continue to increase as investors become even more risk averse. We don’t track recaps and shutdowns in this analysis, but they will also see a pronounced spike. These new trends will probably be true for several years running, until we are solidly into the “Bottoming” phase of the cycle.
There is some good news though! Running alongside the venture capital market cycle is the “Innovation Cycle”, which follows its own S-curve rhythm and is unperturbed by capital market psychology. The innovation cycle is driven by fundamentals of new technology and long-term market needs and continues to spawn new opportunities. Some of the best companies are formed in capital market down-cycles, taking advantage of fresh opportunities and free of market clutter. These founders are usually the hardiest, their teams the most talented and their companies operationally excellent. They are fewer in number, but their impact is often spectacular.