Why Trello Failed to Build a $1B Business
producthabits.com
producthabits.com
In fact, at some point, users started to get anxious that Trello was free and actually WANTED to throw money at the company. I suppose part of that was the perception that if you're paying for it, it will be more likely to survive and won't go away? This was what spurred Trello to create an offering of something (via Trello Gold) - simply so people could pay for something they liked and feel good about themselves, and get something in return.
Actually, I don't know this for sure, but I think they genuinely built Trello because it was a tool that they really felt should exist, and wanted to make it work as inexpensively as they could and support as many as they could, while staying permanently free. I think Trello was an experiment for them, one which succeeded at gaining the huge userbase with a simple tool (which I believe was their goal).
Once the realization that the huge userbase itself was valuable, it was inevitable that someone would buy it out. Atlassian wanted to be associated with that huge userbase, to help with Jira... maybe or maybe not to actually bring Trello into the fold with their existing products.
In any case, Trello is a huge accomplishment, before the buyout, and as a business, even moreso afterwards. I was thinking about building (and actually working on) something like Trello before it existed, but I didn't necessarily think the world needed something so simple. I would have been quite happy with this result!!! They made an absolute ton of users (self included) very happy with the results of their labor, and now they have a wonderful reward from it.
Trello did not fail to build a $1B+ business. Trello succeeded at building a $425M business through good faith and hard work. Bravo.
But not every person wants their business to grow.
"Feed me venture capital, Seymour..."
if the investors got 16% of the company they got $80m, a ~8x return in 3 years.
And more likely they did better. Twitter did a series C, raised $17.4m at $104m valuation, and it was a pretty hot company to give up a bit less than 17%.
A great venture fund gets a unicorn, a legendary one gets two. A 1/2 unicorn that maybe returned most if not the entire fund is a great outcome. If 50% of the investments did that well it's a top-tier fund.
People are a little demanding if only decacorns are seen as wins. It's like expecting your team to win the World Series every year.
Like really, if your advice is useful, why hasn't the author built a $1B business? It seems like they missed the point that Trello became giant for what is essentially networked sticky-notes, were they even trying to become a $1B Business?
Next week: Why Trello Failed To Be the First Company On The Moon (spoiler alert, they weren't trying to do that).
Welcome to clickbait.
> Trello could have created a stickier business product by making sure that it was so deeply integrated with other tools, teams couldn’t rip it out.
But if they had focused on having even more integrations, maybe they wouldn't have been able to focus on creating simple features with broad appeal, or creating a snappy app, or they would have had to spend more money and pivot away from a robust free product.
All of these things would have hurt the user base by ruining one of Trello's main value propositions: it is free and simple. I never would think about learning to use Asana for a personal project, but Trello is an easy sell to teammates for small/medium businesses and personal projects. It is even easy for one-person teams. If Trello wasn't free and easy, then they would just be another Asana competitor, and I can imagine them fizzling with no buyout, let alone an IPO.
I think this sort of business case analysis can be useful, but even in hindsight, the consequences of various decisions are far from obvious.
> They built their servers on top of Node and used MongoDB to store data so that the web app would load really fast.
I stopped reading here.
From a user experience perspective it is freaky fast at updating other clients.
[1]: https://blog.fogcreek.com/we-spent-a-week-making-trello-boar...
It works fine for other boards, so I assume it wasn't built with so many lists+cards in mind. But the fact it, it can be slow, almost unusable, at least for certain use cases.
It depends how much they invested in it before selling, my ignorant guess would way less than $425m. Also some investors may have invested in it for a valuation of 1B, in which case they did not get the return they expected. Anyway, I'm pretty sure all parties involved must be doing fine.
That's the domain of photo sharing apps.
You can talk to it from just about everywhere including the terminal, that's a big reason I use it, it acts as a clearing house for all my stuff, I even use a chrome plugin to store "bookmarks" to stuff I've clipped.
E.g. being able to make layout changes(have lists that span columns for example, or cards that show more of their information on the cover). Or more flexibility in colour changes/background, or ability to style cards.
I've tried many things (including Trello) and even bought a dedicated laser printer and wrote a small shell routine so I could print 3x5 cards to keep on my physical desk. By force of habit and ease of use nothing so far has topped post it notes (other than printed 3x5s or hand written 3x5's) for keeping track of things that I need to keep on top of.
We've been using Trello as the main tool for project/product management for years and quite happy with it. I would gladly pay a reasonable monthly fee for it, but they didn't even try to charge me. I feel like it could've been a bigger business.
Which, I'll point out, is an extremely common scenario. The guys behind Trello stuck to their vision of the product and made an obscene amount of money. Good for them.
How do you keep scaling the revenue? Get bigger customers. Where do you get those big deep-pockets customers? The enterprise. How do you sell to big enterprise? With a big enterprise salesforce.
Ok. With that, you can either build your own enterprise sales machine (which takes years), or you can sell out to someone who already has one. Like, say, Atlassian.
Its not a hard rule that to reach beyond $x you need to become an enterprise
They made it so that one could register easily and use the service quickly, without special training or any setup. That's how you grab tons of free users.
Atlassian is mostly shrink-wrap entreprisey software. They software with expensive licenses that also require extensive installation and maintenance by the client. They need a sale force.
They also host the software. Do you know if the majority of their revenue comes from the self-hosted customers?
Don't let the SaaS thing fool you about sales. Salesforce is SaaS too, but they're a lot more like Atlassian than Google, business-wise.
text is "here is advice to take a burgeoning company to the stratosphere" But honestly, how many people in that position are going to read this blog post? 3? 10? If this was his true intention, he would have packaged this as some kind of super premium offering that only well-connected VCs could read or something.
subtext: "I'm really smart about what takes a company from success to HUUGE SUCCESS." Is this guy trying to gain a rep in VC circles as a CEO of late stage startups or something? What's his angle?
Reaching $N million instead of $2.4N million, where N > 200 doesn't really feel like an obvious shortfall.
Trello Gold with emoji support and custom backgrounds was a weak proposition and they could have moved into enterprise faster. Still, it's amazing they did as well as they did.
Kanban boards and Chat apps are the new todo lists - ten a penny - which makes people dismiss the notion of building a business around such a trivial feature.
This is a mistake. Most of those apps are badly designed, meaning for those looking to eke out a living rather than break $1 billion, there's still plenty of opportunities.
For example, Diigo the highlighting app is relatively unknown but is a hit with educators and so makes cash in a corner where few competitors are looking. The top stopwatch on The Play Store makes over $400,000 with in-app sales.
And Wunderlist sold for $125 million. And Todoist is profitable.
It's a big Internet out there.
Isn't selling a Kanban board for half a billion dollars amazing enough?
Turning a todo-list app into "the single-source of truth for a company" would require monstruous schema changes and changes in all the live updates thing etc., and everything would be slower, heavier etc., all that and you couldn't have known if that would catch up or if you would end up losing your users that liked Trello just because it was a simple board thing. It would be better to start a new app.
I suppose the author built several multi billion dollar businesses?
By any metric, Trello is a huge success.
I mean Heinz sells bottled tomato paste. If pressed, it's probably isn't worth that much to each individual, but they're successful because it's cheap, available everywhere, of sufficient quality, and consistent.
When I moved to an "enterprise" company, I wanted to bring Trello with me, but was immediately told NO because 1) we were already using JIRA and Asana, and a third option was not desired and 2) Trello is (was) not "enterprise-grade" meaning they didn't have sufficient security vetting or centralized user provisioning (and more importantly--deprovisioning.) It was demoralizing and stupid. RIP Trello.
It also is too religiously adherent to the card metaphor (understandable) which works in most contexts but falls down in others where more traditional tools (Jira, Github) shine. We use Trello for a ton, but not mainline work.
I use it as a contact manager, todo list manager, and we use it a lot for card sorting/brainstorming and more "people" type stuff.
Trello didn't build a $1B business because they never forced anyone to pay for their product, even when they were extracting significant value from the platform.
Either way, failing to hit $1bn doesnt sound like much of a failure where I live outside the Valley. $400 some million is pretty respectable.
[1]: https://blog.fogcreek.com/we-spent-a-week-making-trello-boar...
> Hindsight, of course, is 20/20. While I’ve spent most of this post talking about Trello’s missed opportunity, we shouldn’t forget that building a SaaS business that’s worth over $10M, let alone one that’s worth $425M, is a huge accomplishment.
Ignoring the ROI question for Trello investors/employees, I think it's fair to portray Trello as a "failure" if the goal is to build a large self-sustaining independent business, especially if it's true that they "had to sell" rather than stay independent.
RE: Trello, e.g., see the part in the post about "Trello's value proposition". Their $5/month Gold plan was really poorly thought out. Many other areas, again as articulated in the article, we're also not well thought out in terms of building a sustainable, profitable business. Signups != locked in, paying, happy customers.
> The 2013 blog post that announced the launch of Trello Gold highlighted three main reasons users should pay $5/month for Trello:
> Customizable board backgrounds
> 250-megabyte attachments for each card in Trello (vs. 10 megabytes on the free plan)
> Stickers and custom emoji
I can't speak for the author but selling Trello "for only $425m" is a failed outcome if the growth expectations was for it to become a massive $1 billion business.
How do we know that? Because Trello took $10 million in VC money in 2014.[1] So we can just work backward from the $1 billion goal using basic math:
VC firm Index Ventures buys 25% ownership of Trello for $10 million.[2] They manage several funds and I don't know which one they used to write that $10 million check but let's assume the money came from the 6th Fund of $442 million they raised in 2012.[3]
Over the 10-year life of that fund, they basically need to grow that $442 million into ~$3 billion. Since most startups in the fund will turn out to be money losers, they need at least one of their investments in the portfolio to become a $1 billion+ company. And keep in mind they only own 25% which means that even if Trello sold for $1 billion, they'd only cash out their position for $250 million. For Index Ventures to get to ~$3 billion, they need one or a tiny handful of winners from their portfolio to return ~$10 billion! (Assuming they take ~25% ownership in each startup and therefore cash out at ~2.5 billion.)
Yes, IndexVentures still gets ~$106 million (25% of $425m) from the sale to Atlassian but as you can see from the math above, $106m hardly moves the needle toward ~$3 billion. That's a $10 million opportunity cost that they could have put into a different company in 2014.
So, if you're a financial analyst, or one of the Limted Partners in IndexVentures fund, or possibly even J Spolsky, the failure to reach $1+ billion is indeed a "failure."
tldr: because of algebra, if a VC invests in a company, that means they expect the startup to reach a multi-billion market cap.
[1] https://www.crunchbase.com/organization/trello#/entity
[2] https://www.quora.com/At-what-valuation-did-Trello-raise-10M...
[3] http://vator.tv/news/2012-06-17-index-ventures-raises-442m-e...
I put that last "failure" in scare quotes to soften the label a bit because in many ways, Trello is a success.
Nevertheless, the VC will frame it as a "failure" or "suboptimal outcome" because they use the framework of "opportunity cost" which means they missed the other startups they could have invested that 10 million in.
If the VC knows ahead of time that investing $10m startup in will return just 1000% in 2 years, they'd rather ignore Trello and find "another Facebook" that can return 20000%+[1]
[1] https://techcrunch.com/2010/11/22/accel-partners-fund-ix-fac...
Sure, but that's in the context of the fund, not the company. It's meant to be consumed by people dealing with the fund in some way.
> If the VC knows ahead of time that investing $10m startup in will return just 1000% in 2 years, they'd rather ignore Trello and find "another Facebook" that can return 20000%+
No, any fund that knows they will get a 1000% return will definitely invest. The assumption will be that those dollars would have gone to one of the bad investments, because since the vast majority of dollars do, they're likely right. That investment will help the fund, because it doesn't it can be one large return, or a few slightly less large but still great returns. Not every fund that makes money does so because of a single great investment.
There are far, far more metrics, and I'd say far more important metrics to decide whether a company is a success than the amount of VC returns.
Please read my previous comment charitably as I deliberately wrote about Trello using the "investment perspective" for discussion purposes only. It wasn't rendering absolutist judgement that Trello was a failed company.
>So some VCs didn't get the crazy high returns
That's only looking at one side. The other side is J Spolsky who accepted the $10m. To do that, he had to believe Trello would grow beyond $1 billion and convince outside investors of that same vision. If a founder makes a decision to accept VC money, he has redefined the threshold of what success/failure will be. The math makes that unavoidable.
Hitting 3 billion over ten years means getting more than 20 percent growth year on year ten years in a row...
Exactly! When Harvard University invests some of their endowment in a Sequoia VC fund, they are expecting returns better than 20%!
To compare, here's my previous comment about USV's 22%+ returns: https://news.ycombinator.com/item?id=12678009
Only in the waning days of a market top would such an assertion (that such a return is a "failure" to anyone involved) even have a frisson of plausibility.
Also, most of your numbers & ratios are order-of-magnitude right, but as they say, "horseshoes and hand grenades." A $442 M fund size is big enough that returning even a 3x-4x overall, with any consistency, would be downright virtuosic. So, no, they don't need to return $3B or 6-7x the entire fund to be successful (nor do they or their LPs have any realistic expectation of doing so).
The long game, as always, is to try to have just ONE Google in your portfolio (the 400x+). But the short and mid games dominate, and they are about staying alive long enough to have enough at-bats to hit that grand slam. And the absolute, bar-none best way to stay alive that long, is to consistently make money for your investors.
Finally, the ex ante assumptions and hopes and dreams are not really the way to judge the decision to sell Trello. That decision must be judged on facts in hand at the time. And frankly, at that time, Trello was a leading Kanban app, with no obvious or de-risked path toward a 10x or 100x from where they were.
TL;DR: Nobody actually in that game is pooh-poohing a 9-figure exit for 10x in 2 years with a straight face.
It's definitely not failure all around.
>A $442 M fund size is big enough that returning even a 3x-4x overall, with any consistency, would be downright virtuosic.
A 3x return is 11.6%/year which isn't much more than Warren Buffet's Berkshire Hathaway (9.6% last time I checked) -- given the increased risk of VC funds. BH stock also doesn't have overhead of 2/20 fees. Every institutional investor is different but many would look at safer investments if they are targeting ~11%.
If Sequoia, USV, etc are consistently returning 20+%, that's what the competitive levels among top VCs look like. Maybe it's unreasonable for LP's to expect returns like that from a VC fund but they do.
But, the "percent" returns, as opposed to cash-on-cash, aren't actually calculated that way. Most funds eventually call something close-ish to 100% of capital (1.0x paid-in), so you can meaningfully-ish talk about the "times money" on the entire fund. But it doesn't work for "percent" returns, which are necessarily annualized.
Funds call capital over time, not all at once. They also pay back distributions as they become liquid, not all at the end. Investors measure this with IRR, taking into account the timing and magnitudes of the cash flows. So a fund that posts up a 4x "times money" within ~ 10 years is going to have a substantially better IRR than 15%.
WRT LP expectations: institutional investors aren't just weighing VC against "safer investments." They look at long-term historical data about different asset classes, project out their best opinions about likely risk and return AND correlations among classes, and make a top-down decision about where to allocate. Then, within each bucket, they try to select a group of managers who can meet (or ideally safely beat) those assumptions. (If rather small, they may delegate this to a fund-of-funds; if rather large, they may further differentiate into various stages or strategies of venture, and the process is more or less fractal.)
Partly we may be talking past each other because "expect" here could mean various things. One thing that's clear is no major LP / institution is going to be invested only in one VC firm or only one vintage year. They think about these things in aggregate and over time.
If by "expect" you mean, "in aggregate and over time, some LPs project 20% IRR for the venture asset class (subject to a very high, like say 25%+ std dev)" I think that's very fair and you'd be right.
But if by "expect" you mean "LPs look at each venture investment and rely on it to consistently return 20% IRR, or that will have been deemed a failure" then you're not giving them enough credit.
#1 driver of venture performance: fund vintage year.
#2 driver of venture performance: manager persistence (manager skill/alpha).
But to bubble back up to the higher level: LPs who see 10x, 9-figure exits in 2 years coming out of the portfolio of a ~$400M fund are not going to say "why not $3B?" They are going to understand that it's a rare fund in a rare vintage year that is going to return 7x+ on the whole fund, and they are going to be exceedingly happy if the "off years" consistently return 3x.