i'm an early-stage vc - the author's analysis on "number of funds" (specifically VC funds) is accurate. the overall volume of venture allocation has also slowed considerably if not decreased (which is totally expected in a higher interest rate environment).
2021-2022 was a total blip on the screen zero interest rate era thing.
i'm not seeing considerable slowing of new startup development, quite the opposite actually w/ AI. this is for a few reasons:
- accelerators are filling the gap; the accelerator model is actually quite efficient in the early-stage spectrum (it needs some further innovation). there are a huge number of AI accelerators and programs now; and further
- most of the capital going into VC is just being further concentrated into the large Multistage firms like A16Z, Accel, Sequoia, General Catalyst, etc... all of these firms are realizing they need to win deals as early as possible so have multiple seed programs: accelerators, incubations, scouts, fund-of-fund allocation, geographic funds, university focused sub funds, etc...
- overall great founders & startups are truly just exceptional so statistically there just won't ever be that many. venture will always be a cottage industry of sorts. in this form - "venture" equates with "growth"; there can only be 1 category leader by definition and venture is meant to capture this. 2021-2022 overall venture market was too big.
- AI is making startup creation many multiples more efficient. we saw this w/ the advent of the cloud, where startups used to need $2-3M "to buy servers" and 2-3 years to ship a product in 2010, by 2015-2020, they really only needed $3-500k to get a product to market. we're going to see that number come down considerably (unsure if it will be 30-50k, but definitely a lot lower).
- we're also seeing the new wave of the 10-person unicorn (billion $ company); these companies will raise a lot less cash, so will result in higher multiples on the original investment.
- i think the overall distribution of returns will look different on a portfolio basis in 2025-onwards. with power law, we expect to see super long-tail concentration on the 1-2 companies that yield 99% of the return to a portfolio, but i suspect we'll start to see some mitigation of that effect with more companies yielding positive outcomes. this might mean that there's less of a reliance on portfolio construction to generate risk-adjusted returns and that there could be more of a democratization of early-stage investing where we see 10-100x the number of startups and founders. that warrants a longer analysis, but as someone just bullish on startups and everyone being a founder that possibility is very exciting to me.
With so much money flowing into the massive funds, do you think more and more unicorn startups will just stay private? It seems like there are liquidity opportunities for employees/founders via tender offers, secondaries, etc.
If you are a profitable unicorn who can raise money in the private markets when needed, is there really a benefit to going public? Maybe I am missing something, but going public doesn't really seem to be as important as it used to be.
One consideration not mentioned is around developer sophistication. Steve alludes to the expansion effect of CodeGen ("there are millions and maybe billions who are jumping at the chance to code"), but doesn't consider that the vast majority of these people don't know about arrays, data structures, memory, containers, runtimes, etc, etc...
To me, that's the most important consideration here. Are you targeting professional devs who are enhancing their current workflows iteratively with these improvements? Or re-thinking from the ground up, obfuscating most of what we've learned to date?
Maybe we need to trudge through all of these weeds until software creation hits its final, elegant form where "Anyone Can Code".
Maybe the old Gusteau quote is actually fitting here:
"You must be imaginative, strong-hearted. You must try things that may not work, and you must not let anyone define your limits because of where you come from. Your only limit is your soul. What I say is true - anyone can ̶c̶o̶o̶k̶ code... but only the fearless can be great."
Well we'll never reach a state where anyone can code. I have pans, a supermarket nearby, cookbooks and a belly, still I'm never gonna be able to cook, I snooze after 30 minutes, even if I succeed once, I get bored and stop for months etc.
Simplifying to the point a grandma could make an app isn't gonna make any grandma WANT to make apps. And that's fine, there's no issue, we don't have to make more people code and those who want, will, even if all we had was assembly and a light board...
Which I think is the spirit of your quote basically.
That’s a bad comparison, cooking has been done by people for thousands of years, your problem with cooking is laziness, there is nothing mentally or physically stopping your from learning to cook.
I do agree with your second paragraph and it’s more that you DON’T want to cook versus you being unable to cook.
Thats a bad comparison, coding has been done by people for thousands of man-years, your problem with coding is laziness, there is nothing mentally or physically stopping you from learning to code.
I do agree with your second paragraph and it’s more that you DON’T want to code versus you being unable to code.
This is pithy but won't. There's a clear difference between software (no evidence that everyone can do it capably) and cooking (thousands of years of nigh everyone from every background doing so). Laundering it as "thousands of man hours" doesn't change the fact that we've had less than a century of evidence for people coding, and for most of that only a small subsection of the population has picked it up.
Love the movement and glad there are founders out there pushing the envelope for their team.
(aside: 51 points but only 1 comment? It's a front-page worthy article, but sort of feels like there's some vote gaming happening. I've never seen 50 points w/ 1 comment.)
That's actually a lot more common than people assume it is, and comments like "I can't believe there are X points but only Y comments" are more common than you'd think they'd be as well!
My theory is that it's a sign of a good article, because more energy is going into reading it than into posting quick comments (which are usually less valuable comments). But I don't have the data.
Edit: well... we have the data (to test this), but it would be enough of a pain to do the analysis that other things will probably take precedence forever.
I'm one of those people who upvoted without commenting. I think it's just a way of saying, "I found this article interesting / I agree with the content, but I don't have anything of value to add".
The usual problem of late on HN is people commenting without upvoting, even if they like the article.
On this post, I started reading, then paused to hurry back and upvote on HN, to do my part to keep it from falling off the front page, before I returned to finish reading.
I wonder if that's something the algorithm can detect?
Measure the time between when someone clicks a link to the article and when they upvote it, compare that to one of the "estimated reading time" metrics of the linked page...
(Which, of course won't work, because at least some people (ie me) open a bunch of tabs for everything that looks interesting on the homepage, then spend a few minutes at a time over the entire morning choosing a tab and reading/voting...)
I had no intention of posting this to HN (someone I don't know posted it!) and also didn't expect more than the 10 people I shared it with to actually read it - no vote bait intended! I'm happy to take your feedback on a better title that is less baity and more apt.
Are annual plans really that difficult to implement? Doesn't seem like it should be that difficult unless you have complex pricing behaviors. What are some of the most common stumbling blocks with transferring a plan from monthly to annual?
Yearly plans are not the issue. The real problem is to work with different plan intervals: for instance, when a customer switches from a yearly plan paid in advance to a monthly plan, you need to prorate the period they used your product and issue a credit note to refund them for the rest of the period.
Check out https://anycart.com/ - they scrape available delivery slots for you. Also can save you a lot of time / effort and money if you prefer to order by recipe although lots of ingredients are out of stock. (disclosure: investor)
Not surprising to see the common complaints from learn-to-code authors about disrupting their courseware. We heard this a lot at Nitrous.io and struggled with the importance of the beginner market to our business.
When considering Nitrous as a viable business, I'd talk a lot about the "developer sophistication spectrum" and the challenges of one single product or service attempting to meet the needs of a lot of different types of developers.
On the newbie side of the spectrum, serving the hot "learn to code" market means scaling your potential market size by orders of magnitude. There is some product-market fit here as newbies don't really have substitutes ("what's a development environment?" they'd often remark), but the SaaS economics of selling tooling to newbies was atrocious.
Selling to learn-to-code means you're dealing with an incredibly fickle audience where 95% abandon their plans to become a professional programmer within a few months. The other ~5% who become full-time programmers are dedicated enough to their craft to learn about their OS and their options to customize the local development workflow. So they naturally also churn.
(I don't have any knowledge of the market, but I'd imagine courseware providers attempt to charge 100% up-front to account for the extremely high churn. At least, that's how I'd charge.)
So basically all the cloud IDEs are getting hundreds of thousands of signups from a lot of newbies saying "We love [Nitrous, Cloud9, Koding, etc...]!" but not wanting to pay for the infrastructure and churning at unsustainable rates.
On the less sophisticated side of the spectrum, I think there is potential for a viable cloud IDE business, but I think it needs to be closely coupled with a content platform like Treehouse, Coursera or CodeAcademy. I haven't looked at any of them recently, but I wouldn't be surprised if they have in-house teams working to improve the editor experience and provide stateful experiences with dedicated cloud compute & storage. We had a tightly coupled integration with the Flatiron School and it was a pretty solid experience but just wasn't a big enough business for us to scale. So in reality these businesses really just look like a content / courseware business that has a really great cloud development experience. But it's clearly built for people learning to code and they're paying for the courseware, not for the editor.
As you move up the sophistication spectrum, developers begin to experience "cognitive dissonance" when considering how much their time is really worth. That is, when they know how to setup, configure or troubleshoot something themselves, they underestimate the time they spend every month performing those tasks. We spent a ton of time doing deep customer research with excellent engineering teams at Airbnb, LinkedIn, Shopify, etc... You'd be extremely surprised at just how much time it takes for the average developer at a top-tier engineering org -- in some cases, new developers took 3-4 weeks to setup their dev environment. But after setting up a new environment the other dev ops problems start to spider into a web of complex and proprietary issues that are difficult to create compelling marketing / sales presentations. It's like - everyone knows it sucks and it's broken, but nobody quite knows the solution. Which is why a lot of the solutions emerge from open source projects that solve specific issues organically and then expand into powerful platforms that cohesively solve a set of interesting ops problems (e.g. Hashicorp).
This is an oversimplification of the complexities of the developer market - as there is also a spectrum of sophistication within the professional developer market itself. The "intermediate" professional developer tends to be the best market fit right now for cloud development / IDEs, as they often are self-taught and know how to code, but are often not as versed in debugging low-level issues, but usually are more price sensitive to their more sophisticated counterparts (who don't want to use the service in the first place).
In any case, I remember reading a HN comment about the nitrous.io shutdown [1] and feeling bad about not opening up more so I suppose this will provide some color. People loved our service and we honestly loved building it, but business is hard and we weren't able to uncover the right strategic focus. Hopefully Coursera, Treehouse, CodeAcademy, etc... will continue to fill in the gaps for the beginner market - but since those will be tightly coupled with their courseware, it's going to be a difficult spot to be in for the independent educator who is attempting to monetize their own material.
There's a lot of really interesting things in this post, but I'm only going to pick up one point:
"As you move up the sophistication spectrum, developers begin to experience "cognitive dissonance" when considering how much their time is really worth."
I am pretty sure that one of the reasons why cloud IDEs have struggled is that plenty of professional developers consider that learning an IDE or editor is a time investment that pays returns over a long period of time. On the extreme end of the scale, I still get near-daily use from the Vim knowledge that I picked up well over a decade ago. Investing personal time learning a cloud IDE may not be such a good investment, because those skills can be invalidated at any time.
There's definitely a lot of potential innovation that can be done in developer collaboration and environment setup, but I don't think that a proprietary system will get the widespread adoption needed to move the industry forward.
It’s easy to forget that good ideas take a lot of time and iterations until they become magical. Great things do not happen overnight and without much sacrifice. The majority of people would give up the first time the community dismissed their prototype.
Exactly my thought. You need a bit of luck along with good project to get viral on HN. Otherday I saw simple PDF memo generator got lots of traction while so many other cool apps didn't even surface up.
We're working on a static site hosting service called PubStorm (https://www.pubstorm.com) that allows SSL on custom domains. It's free; we're also working on some powerful features that will set it apart from gh-pages.
Entrepreneurship is more about conviction, determination, and will than ideas and market opportunities.
To me, this letter is basically saying you're just not up for the challenge. That's fine -- it's certainly better to admit it now, and it's likely a rational decision.
But the best companies usually aren't created by rational thinkers. The best entrepreneurs can take a horribly broken system and contort it in ways that nobody else even thought of, and then execute on a vision to make it the new reality.
You're not telling your investors "this market is too tough, let's not go here", you're basically just saying "This market really needs to be un-fucked, but I'm not the guy to do it".
I believe your generalizations go too far. The "crazy" founder who succeeds despite all odds is only one archetype. There are others. Many outliers are successful by creating new markets.
This guy (T.A. McCann) has successfully built and sold a startup, so he doesn't lack the will -- he just lacks the will for this business.
Markets can be so tough that the time/cash/energy to succeed in them means the opportunity is not as attractive as others.
Actually entrepreneurship is ALL about market opportunities and ideas. All you do is maximize all the odds and leverage everything you have for a win. Winning is all that matters, and winning as big as you can.
Now if you want to take on a big challenge, change the world, and build something magical, you can and you should. However none of these are required, and each of these is more likely to be a problem.
2021-2022 was a total blip on the screen zero interest rate era thing.
i'm not seeing considerable slowing of new startup development, quite the opposite actually w/ AI. this is for a few reasons:
- accelerators are filling the gap; the accelerator model is actually quite efficient in the early-stage spectrum (it needs some further innovation). there are a huge number of AI accelerators and programs now; and further
- most of the capital going into VC is just being further concentrated into the large Multistage firms like A16Z, Accel, Sequoia, General Catalyst, etc... all of these firms are realizing they need to win deals as early as possible so have multiple seed programs: accelerators, incubations, scouts, fund-of-fund allocation, geographic funds, university focused sub funds, etc...
- overall great founders & startups are truly just exceptional so statistically there just won't ever be that many. venture will always be a cottage industry of sorts. in this form - "venture" equates with "growth"; there can only be 1 category leader by definition and venture is meant to capture this. 2021-2022 overall venture market was too big.
- AI is making startup creation many multiples more efficient. we saw this w/ the advent of the cloud, where startups used to need $2-3M "to buy servers" and 2-3 years to ship a product in 2010, by 2015-2020, they really only needed $3-500k to get a product to market. we're going to see that number come down considerably (unsure if it will be 30-50k, but definitely a lot lower).
- we're also seeing the new wave of the 10-person unicorn (billion $ company); these companies will raise a lot less cash, so will result in higher multiples on the original investment.
- i think the overall distribution of returns will look different on a portfolio basis in 2025-onwards. with power law, we expect to see super long-tail concentration on the 1-2 companies that yield 99% of the return to a portfolio, but i suspect we'll start to see some mitigation of that effect with more companies yielding positive outcomes. this might mean that there's less of a reliance on portfolio construction to generate risk-adjusted returns and that there could be more of a democratization of early-stage investing where we see 10-100x the number of startups and founders. that warrants a longer analysis, but as someone just bullish on startups and everyone being a founder that possibility is very exciting to me.