VCs Fund Ideas, Not Execution
My founder journey started in 2014. I have pitched to hundreds of investors and raised 15M for my last startup, taking it from 0 to $1M ARR. My startup was in the graph database field. And I frequently wondered why other database companies were able to raise faster and more than us. What was it about them that made them special: “Why them? Why not us?”
I’ve learned a few things that I think would have been incredibly helpful to me before I started my company. As I write this, I think about my past self and all the future founders who could gain from this hard-earned wisdom.
People tend to think
Ideas are a dime a dozen. Execution is all that matters.
This saying is so wrong, that it should be banned. If you are fundraising, ideas matter way more than execution. Ideas dictate which market you’re playing in, which dictates what your company could be worth.
A lot of founders, myself included, who believe in execution over ideas spend way too much time perfecting their execution. They should instead be spending that time tweaking their idea until it falls in the right market. Which market they pick gives them an inherent fundraising accelerator and multiplier, which directly impacts how fast they raise and how much they raise.
Startups don’t need more traction to be worth more
To illustrate this point, I analyzed 10 database companies — 5 in the SQL market and 5 in Graph/Multi-model DB market. The companies I tracked are popular in their respective markets. I tracked their fundraising over time and their annual recurring revenue 1.
What you’d expect to see from this analysis is that startups with better revenue traction would be able to fundraise more — particularly on and after Series B. As the data shows, this is not how funding worked though.
A founder who started a SQL database company would raise $200M within the first 5 years of their journey from top Silicon Valley investors. On the other hand, a founder in the Graph/Multi-model database market would take double that time, i.e. 10 years, raising from an assortment of investors across the world if they even reach there at all.
Out of the 5 Graph/Multi-model companies I investigated, only 2 raised more than 200M USD, hitting that around the 11-year mark. Compare that to 4 out of 5 SQL companies, who all raised over 200M USD within 5 years. The one which hasn’t yet (PlanetScale) is only 4 years old but having raised over 100M by year 3, is well on its path (and probably would raise a round hitting $200M in year 5).
In fact, none of the “Tier 1” Silicon Valley VCs ever invested in any Graph/Multi-model DB company despite the companies/founders being located in the valley. They instead raised from European, Chinese, or other US (non-Valley) investors. On the other hand, all SQL companies raised exclusively from the valley, from Tier 1 investors, with AirTable being the first one to go outside the valley, raising from New York (Thrive Capital) in their 5th year 2.
Compared to its SQL peers, AirTable seems to be on a slightly delayed trajectory in hitting the $200M mark compared, but still beat out all the Graph/Multi-model databases I investigated by having raised over $150M by its 5th year.
SQL market is huge, So revenues must be too
Alright, SQL startups get more funding. That’s okay because SQL is a huge, well-proven market compared to the Graph/Multi-model market. A bigger market should generate bigger revenues. You’d expect to see the SQL companies outperforming their Graph/Multi-model peers in terms of revenue.
But that’s not what I saw. The revenues were comparable, if not sometimes more. Arango and Yugabyte (SQL) are estimated at 10M ARR. Fauna, Cockroach (SQL), and TigerGraph are estimated at 15-25M ARR. Neo4j and Airtable (SQL) are estimated at 100M ARR.
When comparing the money raised against estimated revenue 3, you see a big factor difference between SQL companies and Graph/Multi-model companies. SQL companies raised at 20-30x (with AirTable “lagging” at 14x), while all Graph/Multi-model companies raised at 4-7x.
The bigger SQL market didn’t necessarily translate into significantly bigger revenues. Which leads to the question — What are VCs looking for in an investment? Why do they invest?
There are 3 reasons why VCs invest
Investors fund markets, not companies.
Conviction is gut-feel and pattern matching. It’s simultaneously the cause of questionable fundraises and gender disparity, while also being the real reason why Silicon Valley works in the first place.
“Stanford grads have done great. You’re a Stanford grad. So you’d do great.” — Pattern matching
The conviction could be around the people. They might like the founding team, because the founders went to a prestigious school, worked at prestigious companies, happen to talk smoothly, or a buddy of a buddy had worked with the founder in the past and liked them.
Most likely though, the investor has conviction in the market the startup is playing in. The market has seen multiple successful exits, so doing something/anything incremental in the market is bound to get traction and could get the returns.
Or, this could be a radical new technology proven at a big company (Spanner, Kubernetes, GraphQL). The fact that the technology was already built and deployed proves that it is possible, which de-risks the technology, particularly if you can get the original authors / a capable team involved. A BigCo behind the tech also makes marketing easier.
“This tech is powering BigCo. You can trust it.”
“If it works for BigCo, it should work for you.”
Those are very convincing arguments to sell to enterprises. Hence, it is easy to get conviction around funding a startup bringing the BigCo tech out to the masses.
Conversely, a mixture of “radical” tech and “unproven” leads to a negative conviction. The graph database market suffers from this. No BigCo has built and deployed a brand new graph system from scratch (Google and Facebook used existing stacks to build their graph offerings) — which means there’s little conviction around the tech.
Traction is one of these two things. Revenue (recurring / growth / NDR, etc.), or number of users (active users / growth / engagement). Traction is the true north star metric to judge a company.
In a perfect world, where conviction played little role, traction-based investment would lead to meritocracy — a technology would be able to stand purely based on how well it’s received in the market, how many users it has gained, or how much revenue it’s generating.
But in reality, traction is seen from the lens of conviction.
Series B startups and onwards should be judged purely based on their traction. However, this is not the case with database startups — which continue to get a multiple on their revenue based on which market they fall in (conviction). Hence proving that conviction continues to play a pretty big role even in later stages, even when traction should take over (money raised vs. est. revenue chart above).
3/3. Fear of Missing Out
There could be a deal-level FOMO. Here I’m talking specifically about FOMO at the market level. This typically applies to “hot” markets — it could be the AI/ML market, or Crypto market — something with a strong tailwind behind it. Markets that are so hot that it would be more questionable not to invest in it than to try to build a conviction or ask for traction.
What about execution?
Product-focused founders (myself included) might think that building a better product, more traction would lead to more fundraising. A more complex software surely needs more funding than an easier one.
In reality, all those factors are ignored.
A better execution should lead to a better product. But, product quality is subjective. It is hard to objectively differentiate a good product from a bad product, even for engineers, leave aside financiers / MBAs. So, traction becomes a proxy for execution.
“If your product is so good, surely your users would pay you for it.”
Users care about the product (VCs don’t or can’t). They can tell a good product from a bad one. And they would pay for it. But, as we saw before, even better revenue doesn’t mean a bigger fundraise. The multipliers still exist — set in stone purely based on market conviction.
And because those multipliers exist, the execution argument becomes more complicated. Money attracts talent. If a startup can raise more money, it can put together a “more talented” team (whether they are more effective or not is a separate question). Big fundraising announcements and a popular team generates more marketing, making more potential buyers aware of the product, which should lead to more traction.
Great team meets lousy market
As per Andy Rachleff (link)
“When a great team meets a lousy market, market wins. When a lousy team meets a great market, market wins. When a great team meets a great market, something special happens.”
While this quote seems so on-point, it probably isn’t. My analysis shows that it is not that the market is lousy, it’s the “conviction in the market” that’s lousy.
The Graph/Multi-model databases performed just as well as their SQL counterparts in terms of revenue. And yet, the funding continued to stay a factor lower.
It’s also hard to assess what makes a “great team”. It’s as subjective as a “great product”. A proxy for a great team (and the resulting great product) would have to be traction. And SQL companies I analyzed that got more funding didn’t necessarily end up with more traction than their Graph/Multi-model counterparts.
If you don’t care about fundraising
Just focus on the product and traction. All the charts indicate that you don’t need to raise more money to achieve better revenue/users (traction). And for a bootstrapped startup, traction should be the north star metric to focus on.
But if you are in the fundraising game, game Conviction
VC funding is not a traction-based meritocracy. It is a pattern-matching-based gut feel. Strolling in with an idea in the right market (conviction) into a VC pitch meeting is better than convincing them of better execution (traction).
VCs are diversified. Founders have one shot. Choose your idea wisely.
Follow-up on Common Critique of this article
Some friends of mine raised interesting critiques about this article.
- But Manish, surely the SQL market is bigger. Look at Oracle, they’ve done incredible. I don’t see anything in the graph market which is as good.
And you’re right. Oracle and Snowflake (see footnote) both help build conviction in the market. Because Oracle did well, other SQL companies are expected to do well (pattern-matched conviction). It doesn’t mean any individual database company is doing well (traction).
Even if a non-SQL startup has better traction, it still can’t hope to replace Oracle. And therefore, would continue to get fewer funds.
And that only proves the point I’m making in the article. The idea conceived at the birth of the startup dictates how much you raise and how fast you raise, based on the conviction that’s already built in the past around the market.
- It’s not just the market, it’s also timing.
Yes. The idea is a proxy word for various things which once set can’t be changed by a founder irrespective of how much dedication or brilliance of execution they put in.
- It’s also about the team — Investors invest in exceptional teams.
It surely is. If you had multiple plays in a “hot” market, you choose the founders who can win in the market. But, the set of founders to consider would still be filtered by the market they’re playing in.
Conversely, if there’s little conviction in the market, exceptional founders can’t change it without having created a wonderful exit first — A catch-22 situation.
Catch-22s often result from rules, regulations, or procedures that an individual is subject to, but has no control over, because to fight the rule is to accept it.
TL;DR When it comes to fundraising: Conviction > Traction. Idea > Execution.
P.S. An investor who I reached out to for comments on this article mentioned that due to recent public market correction, the Series B and onwards funding is tighter now than before — depending more on traction than conviction. But, I couldn’t find any references to verify that.
The raises in the charts are tagged with the location of the investment company. SV = Silicon Valley. NY = New York. US = Rest of the US. EU = Europe. CN = China.↩︎
Snowflake and Couchbase omitted from money raised vs. revenue, as they are public companies now. As of publishing date, Snowflake is worth over 60B, while Couchbase is worth under 1B in public markets. Their public performance would lead to more (pattern matching and) conviction, hence enabling or hindering fundraises for the startups in their corresponding markets.↩︎