Product Market Fit
Published September 30, 2024
If your company cannot withstand a competitor giving away your product FOR FREE, you do not have product market fit.
This doc is an internal Motion doc that was presented at the company’s engineering all hands on October 3rd, 2024. It has been very lightly edited to release to the public.
I am talking about this today because I felt that what leadership thought about the state of the business and what ICs thought had diverged significantly enough to merit a conversation.
This is typical of most companies. This kind of re-alignment is common at Motion every six months or so. This is the first that’s been substantial enough to warrant a document and a recording.
We Are NOT at PMF
In 2005, an ecommerce brand that focused on new parents (diapers, soaps, mommy kits) burst onto the scene. At the time, it was the fastest growing ecommerce brand in the US. They did this in large part my modeling their operations after Bezos — going so far as to call him “sensei” in their private chats.
Blackburn ominously informed the Quidsi cofounders over an introductory lunch that the e-commerce giant was getting ready to invest in the category and that the startup should think seriously about selling to Amazon. Lore and Bharara replied that they wanted to remain private and build an independent company. Blackburn told the Quidsi founders that they should call him if they ever reconsidered.
Soon after, Quidsi noticed Amazon dropping prices up to 30 percent on diapers and other baby products. As an experiment, Quidsi execs manipulated their prices and then watched as Amazon’s website changed its prices accordingly. Amazon’s famous pricing bots were lasered in on Diapers.com.
[sic]
September 2010, Lore and Bharara traveled to Seattle to pitch Jeff Bezos on acquiring Quidsi. While they were in that early-morning meeting with Bezos, Amazon sent out a press release introducing a new service called Amazon Mom. It was a sweet deal for new parents: they could get up to a year’s worth of free two-day Prime shipping (a program that usually cost $79 to join), and there was a wealth of other perks available, including an additional 30 percent off the already-discounted diapers, if they signed up for regular monthly deliveries of diapers as part of a service called Subscribe and Save.
Amazon was on track to lose $100 million over three months in the diapers category alone.
This is what true scorched earth looks like. Amazon would rather lose ~$400 million a year on diapers alone than let you win. And during the negotiations when you’re waving the white flag, Amazon continues to twist the knife in deeper. A press release that’s going out literally as you’re talking to Bezos about pricing. And this is how Bezos will treat someone who respects the heck out of him — imagine what he’ll do to you. Amazon has done this exact playbook half a dozen times by the way. Maybe the most famous example is Zappos:
In early 2007, with apparel brands watching closely for any signs of discounting, Amazon added a five-dollar bonus to its free overnight shipping. In other words, a customer was given five dollars just to buy something on the site. It was a clever but transparent ploy, an effort to inflict further pain on Zappos. Employees who worked on Endless say that, naturally, this was Jeff Bezos’s idea.
[sic]
[Michael Moritz] had watched Amazon destroy one of his portfolio companies, eToys, a decade earlier and knew that to compete with Amazon, Zappos needed more engineers and more sophisticated fulfillment capabilities.
To put this in context, one of Zappos’ innovations was the way it could deliver shoes so fast. It was a new warehousing strategy. And Amazon didn’t want to lower prices lest it scare away 3rd party retailers. So it just gave a targeted bonus only to ‘overnight shipping’ of $5. Why would anyone object? Any customers of third party retailers would just get their order faster. No third party retailer offered their own shipping… other than Zappos.
Lest you think this is just an Amazon problem, let’s look at how Facebook operates regarding the threat of Instagram.
I think it’s a really big deal that we ship this photos app quickly. [sic] I view this as a big strategic risk for us if we don’t completely own the photos space. If Instagram continues to kick ass on mobile or if Google buys them, then over the next few years they could easily add pieces of their service that copy what we’re doing now, and if they have a growing number of people’s photos then that’s a real issue for us.
They’re growing extremely quickly right now. It seems like they double every couple of months or so, and their base is already ~5-10m users. As soon as we launch a compelling product a lot of people will use ours more and future Instagram users will find no reason to use them. But at the current rate, literally every couple of months that we waste translates to a double in their growth and a harder position for us to work our way out of.
I get that there are also team issues and we’ll need to work through those, but if there are actually people available who could build this more quickly then I think we’d be crazy not to do that. [sic] The message I got from your note though was that you care more about fixing the team than shipping the app. I think we need to do both, but I do think it’s a crisis that we don’t have a mobile photos app out and I’d prioritize pushing that out as much as possible. Getting the team to a good state is not a milestone by itself that I care about.
Separately, I just don’t understand why this app is so hard to build. It basically seems like a camera view, some filters, controls for deleting/tagging/captioning your photos, integration into a composer that already exists, and a skin for a photos feed whose infrastructure should already exist. We’ve had designs that I’ve been fine with for months. The only piece that seems technically new for us is the filters, but it seems like we’re doing okay on those now after a long parallel track. What is so hard about the camera view, feed filters and composer integration? Are we building other unnecessary features?
I guess my basic point is this: I get that your team has issues, so fix them and ship the app. I don’t see why this should be so hard, and I don’t think we should accept any excuses for not getting this done in a short period of time.
— Mark Zuckerberg, in emails revealed during a lawsuit
The above was from an email sent in 2011 when Facebook was worth 50 billion dollars. Most people would be rather complacent by this point. Zuck sets up a clone of Instagram and takes it to the negotiation to the Instagram founders! And by the way, it totally works. The fear of this instagram clone (aka Mark going “destroy mode”) is what ultimately compels them to take the offer - and ultimately what saved Facebook over the next decade as Facebook.com itself began declining. It is not an exaggeration to say that it saved Facebook.
All things considered, Instagram made out pretty well. They had very few employees, no real prospects of revenue, and Zuck still handed them the largest social media acquisition in history (at that point). Things could have been much, much worse. This stark retelling by Andy Miller negotiating with Steve Jobs is what negotiations usually look like when you don’t have product market fit.
And he looks at me, and says, “How much am I buying your company for?”
“We have a deal. 325 million.”
And he looks over at Peter Oppenheimer who is the CFO of Apple, and he’s like, “What? I’m spending 325 million for Andy’s piece of shit company?”
And Peter said, “Well that’s what we’ve negotiated.”
And he goes, “Well I’m not paying that, Andy.”
“Well that’s the deal here.”
“Well I’m not paying that. Your company’s not worth that.”
“Well, I think we are.”
“So, what can we do?”
“Uhh, I don’t really understand. Are you asking me to negotiate against myself?”
“I’ll pay you 275.”
And I say, “Uh, no we have a deal man.”
“We don’t have a deal until I say we have a deal, you understand?”
“Fair, but I think you’re re-trading.”
And he says, “Are you calling me a re-trader?”
“With all due respect, I think so - cuz I came here thinking we have a deal, and I told my board, we have a deal for 325, and I’m going to walk out of here at 275?”
And he leans in close, and he says this. “How about you tell your board back in Boston that ‘oops, Andy’s little ads don’t work on Apple phones!’ And then see how good your business is and how much that’s worth?”
By the way, this is Steve Jobs on his death bead. He’s 100 pounds, and earlier in the conversation Jobs tears up saying he’s not sure if he’ll be around to see his kids attend college. This is how Jobs negotiates on his deathbed.
This is Motion’s future if we don’t find product market fit. Multiples, free cash flow, future growth — none of these matter. If we don’t find product market fit, then we’re worth whatever our aquirers say we’re worth.
In case these stories aren’t getting the message across, let me be blunt: It is only by the grace of God that we do not have a competitor eyeing to take our lunch right now. And the moment the Eye of Sauron is upon us, we must be ready to withstand scorched Earth.
I suspect most of you have never had to come face to face with this. Well I have. People forget just how ruthless the early days of Uber vs Lyft were, and let me tell you the scooter wars were no different. Ex-Uber employees were all over Lime, Bird and even my company, Skip Scooters. So the “Uber playbook” which was once revolutionary was now the norm. We did everything within the limits of the law to take down our competitors (and often times in the gray zones of the law, too). Trash their scooters. Relocate scooters to illegal areas and send city officials “anonymous reports” to get licenses revoked. Reverse engineer their apps to see if we could just DDOS them. Whatever you can think of, we did it. I’ve seen it first hand, and yes, I’ve even implemented it myself.
Monday has 300 engineers. Notion has 500. Atlassian has over 1,000. How long would we last if any one of them decided to build a sub-par Motion clone and just give it away for free? How long would we last if any of them decided to try and do everything in their power to simply bring us down?
So when Harry says that Motion has not yet found product market fit, maybe you understand a bit of where he’s coming from now. We will not be announcing our metrics or our fundraising externally. We will not be bragging about how well we’re doing to assuage our own ego. And we will not be deluding ourselves into thinking we have actually achieved something. Because we haven’t. Not yet.
If you want to see what PMF looks like, here’s how Snapchat responded to Facebook launching an exact clone. Keep in mind this takes place just 8 months after Zuck purchases Instagram.
Zuckerberg asked probing questions about Snapchat and their vision for the product and company. He then wondered aloud what Snapchat would look like as a Facebook-owned company, with Evan and Bobby still at the helm, able to take advantage of the social giant’s resources and funding to grow more quickly — as Instagram had. And indeed, Zuckerberg had an impressive story to tell. Following its acquisition, Instagram’s daily active users grew almost 1200% in just six months. Perhaps Facebook would be interested in acquiring Snapchat for $60 million, instantly making Spiegel and Murphy millionaries in their early 20s.
Evan explained that they weren’t interested in selling the company.
In response, Zuckerberg showed them something new that his team had been working on: Poke, a new Facebook app, would be released in a few days. What was it? A messaging app for disappearing photos and videos. The message was clear. Join us, or we will crush you.
[sic]
On December 21st, Evan received a one sentence email from Zuckerberg. “I hope you enjoy Poke.” Evan had deactivated his Facebook account, so he couldn’t even access the app. In a panic, he called Bobby who downloaded Poke and made an account. It was an exact replica of Snapchat. An unabashed copy. Poke had even stolen Snapchat’s user interface for recording video. And in addition to sending disappearing photos and videos, users could even send disappearing text messages and even just poke each other to get attention, like in the earliest days of Facebook when it was fun and weird. But there was nothing fun about Poke. This was a show of might from Zuckerberg and his team up north.
Veteran Facebook director of product Blake Ross led a team to develop Poke with a small team in just 12 days. Zuckerberg, who invented poking years prior, even wrote some of the code for Poke, even though he rarely programmed at Facebook anymore.
— How to Turn Down a Billion Dollars
And what happened was later described by Evan Spiegel as “the greatest Christmas present we ever had”. Facebook’s attempts to kill Snapchat was the very thing that helped it grow — all the way to the top of the App Store just a week later. That is strong product market fit.
Velocity Is Everything
There’s a natural tension between engineering and product as companies grow. Engineering is the bulwark against bugs while business is the bulwark against complacency. But I present the business case above so that, as engineers who are likely less exposed to the realities of the business world, you can always keep this in the back of your mind. The motivations from folks on the business side are fundamentally good. Let’s not become the next Zappos or Diapers.com. Tech debt is not a moral failing, and it’s not going to kill us. Our inability or unwillingness to deal with it, however, definitely will.
So how do we win? We’re clearly outmatched and swinging against big sluggers. The answer, as it so often is in business, is velocity. We’ve been blessed with a head start, and it’s our job to maximize it. Do not expect it to last forever. It never does.
As soon as we hatched that plan for ourselves, it became immediately obvious that every company in the world was going to want this. What really surprised us was that thousands of developers flocked to these APIs without much promotion or fanfare from Amazon. And then a business miracle that never happens happened—the greatest piece of business luck in the history of business, so far as I know. We faced no like-minded competition for seven years. It’s unbelievable. When I launched Amazon.com in 1995, Barnes & Noble then launched Barnesandnoble.com and entered the market two years later in 1997. Two years later is very typical if you invent something new. We launched Kindle; Barnes & Noble launched Nook two years later. We launched Echo; Google launched Google Home two years later. When you pioneer, if you’re lucky, you get a two-year head start. Nobody gets a seven-year head start, and so that was unbelievable.
— Invent and Wander: The Collected Writings of Jeff Bezos
Those of you who came to book club will already know this, but Bezos was so intent on keeping AWS financials a secret that he shoved it a nondescript “other” category for the longest time. Amazon finally disclosed AWS financials for the first time in 2015, 9 years after it was created, mostly because of an SEC ruling — any line item that is more than 25% of revenue must be independently reported on. Otherwise, I’m confident that to this day we wouldn’t know how much AWS really makes. When you have a winner, you don’t brag about it. You shut up and ruthlessly execute.
Over its first decade, AWS’s revenues and profits were a closely guarded secret. The division generated $4.6 billion in sales in 2014 and was growing at a 50 percent annual clip. But Amazon disguised those numbers, along with nascent advertising revenues, in a sundry “other” category on its income statement, so that potential competitors like Microsoft and Google would not recognize how attractive a business cloud computing actually was.
— Amazon Unbound by Brad Stone
Most people think Motion is at 1-5 million ARR. They laugh or give me pitying smiles when I tell them I work at Motion. That’s okay. Embrace it. A willingness to be misunderstood is crucial in any great enterprise - we are starting with a contrarian premise. Most people will disagree. It will be preposterous until it’s obvious. We will continue to stay secretive to build on our premise.
It’s good that they laugh. Their laughter is what is giving us this incredible lead on the competition. Unless we’re supremely lucky, in under 2 years competitors will start paying attention to Motion as well. So it’s vital that build on this advantage as much as we can during this time.
Many of you thought that after PMv3 and hacker house3, the pace of our execution would slow down. I’m here to tell you that the exact opposite is going to happen. We’re going to continue executing at an incredible pace and look for ways to get even more velocity so that we can compound on our advantages. This is why I’m so obsessed with shifting left, making CI faster, and finding ways to hacker-house feature branches and get them in front of customers. By the time Notion or Monday start thinking about copying the features we’re working on now, we’re three or four features ahead. Ideally, we’re so far ahead that we dissuade them from even competing.
After all, as Peter Thiel often says, competition is for losers.
Iteration velocity solves all known software problems. Eric Schmitt at Google always said “revenue solves all known problems”. I think that’s actually bullshit. Revenue doesn’t solve Google’s anti-trust problems. But if you’re a professional software engineer, you have to realize at some point, that you don’t know the future. You’re going to make a wrong turn here or there. But if you are able to react quickly, then it’s not as bad. You can adjust quickly.
Our job right now is not to polish every diamond and make it bullet proof. I want to make that very clear. Our job is to make the system extensible and modifiable so that we can iterate. Everything we focus on from a tech debt side, infra side, and feature side should be focused on increasing our iteration speed. Expect that we will iterate. And banish the notion from your mind that what you’re building now will last for another three years. If we do our job and find product market fit, it will not last. We will do another rewrite.
Understand that we’re in the risk taking business. Because we don’t have product market fit, because we are not an incumbent, we cannot afford to not take risk. If we do something after everyone else agrees it’s a best practice, then we’re already too late. Product and business’ job is to push us to do things that we may not want to do.
So when it comes to new features, we will always be taking on some amount of tech debt and risk. If you polish a feature until it’s done “right” and there’s absolutely no tech debt, you probably spent too long on it.
There’s an attitude starting to take place where certain folks are ultimate gate keepers with veto power, and other people have to go to them and beg them to do stuff. Bezos calls this phenomenon the “institutional no”. There’s an incredible unwillingness to do things — and these are not ill intentioned people! They usually have the best of intentions, and often, they’re even right about the things they’re objecting to!
This dynamic arises because in many companies, one department is, indirectly, responsible for the actions of another. Imagine a security team who is ultimately on the line for every security vulnerability. Of course, they are not the ones developing new features. So what inevitably happens is every feature now must go through sec-review, and the security team either prevents the feature from shipping or gets overruled by a director, leaving them frustrated.
This sort of phenomenon happens all the time, by the way.
A few weeks into these discussions, in an S Team meeting, Bezos announced that Amazon would develop its own dedicated electronic reading device for long-form reading. It was a stunning edict. Creating hardware was expensive and complicated. It was also well outside of Amazon’s core competency—its litany of obvious skills. There was a chorus of vehement objections. Jeff Wilke in particular had the background in manufacturing to know what challenges lay ahead for the company if it tried to make and sell its own devices. “I thought it would be difficult and disruptive and I was skeptical that it was the right use of our resources,” he says. “It turned out that most of the things I predicted would happen actually happened, and we still powered through it because Jeff is not deterred by short-term setbacks.”
— The Everything Store (emphasis mine)
Just because a particular path will cause some problems (like in the case of Kindle), doesn’t mean that path is wrong. It just means we have to be ready to deal with it when the time comes. Similarly, just because someone knowingly added tech debt or created a mess to clean up does not make it a bad call. Moving with urgency - in any direction - is better than standing still. Make a decision and move on. Almost everything you’re arguing about is a type 2 decision that can be reversed later on.
Obviously, stability plays a larger part the closer we get to PMF. Unlike our first S curve2 with consumer, we have to maintain some floor of stability while going just as fast as the first time. These are conversations I’m very interested in (how can we go fast in a more stable manner) as opposed to the naysaying “we need to slow down”. Because the truth is that we want to slow down, but we need to go fast.
Enabling speed with stability is my core responsibility at Motion. It’s the core criteria I look to when evaluating tech debt projects. It’s also, ultimately, the core mandate of Sean’s pod. Stability for its own sake is not something I nor anyone at this company is interested in. We are not NASA. We are not a database. And we are not a medical device company. If we were, we would operate in a very different model. By our very nature, by our industry, we are allowed to operate with more risk. If we don’t take advantage of that, then we take on all the downsides of the velocity hit without any upside (regulatory moat, marketing positioning, etc). Instead of wishing Motion was the tortoise, ask how can I make this hare survive a marathon.
Making a decision - any decision - is often the most useful thing you can do. Move the company forward, constantly.
You should read biographies about CEOs, about business. And you would see that even Steve Jobs, he made probably 90% of mistakes in all his decisions. And many, many stupid mistakes. And he was not good altogether. Maybe the single common trait between all successful CEOs I may found is decisiveness. You should make decisions. And because when you make decision, you move forward. Even if decisions are stupid, you are moving forward and you are getting new information and next decision will be better. The worst is just this analysis paralysis not making any decisions.
— Dimitry Gurski, CEO of Flow Health (emphasis mine)4
Remember how long we argued about the activity feed? Reverse chron was supposed to be terrible, we spend days going back and forth. We ended up doing two more iterations on it - and let me tell you - I’ve personally had no fewer than 6 messages from B2B customers thanking me about how great it is.
The fact is, v1 and even v2 of activity feed were bad. But we will iterate until we get there. Don’t argue against trying something because you think it might be a mistake. Embrace that we’re going to make mistakes. Make them anyway. Anything we’re proposing is not the end finished product. It’s the starting point.
Failures Are a Feature, not a Bug
As a company grows, everything needs to scale, including the size of your failed experiments. If the size of your failures isn’t growing, you’re not going to be inventing at a size that can actually move the needle. Amazon will be experimenting at the right scale for a company of our size if we occasionally have multibillion-dollar failures. Of course, we won’t undertake such experiments cavalierly. We will work hard to make them good bets, but not all good bets will ultimately pay out. This kind of large-scale risk taking is part of the service we as a large company can provide to our customers and to society. The good news for shareowners is that a single big winning bet can more than cover the cost of many losers.
— Invent and Wander: The Collected Writings of Jeff Bezos
There’s this really incredible Youtube video by Bain & Company that I recommend everyone watch. Long before Paul Graham’s founder mode took over the internet, Bain talk about “Founder Mentality”. In the early days, companies are basically like insurgents. You don’t have a lot to lose, and so you frequently go all in on various propositions. But as you grow, you adopt an incumbent mindset. Incumbents are also willing to try a lot of things, actually. The key differentiator is that incumbents lack the willingness to go all in. What if it blows up? What if it’s embarrassing? What if I lose the wealth that took me over a decade to accumulate? And so the size of your failures reduces.
It’s one of the reasons I’m so bullish on Zuck. Not only is he one of the youngest founders still in the game, but time after time, he’s shown he still has it. There’s an enormous disconnect between people who understand this principle — that in fact big failures are a feature, not a bug — and those who don’t.
These kinds of tweets, while quite funny, are a great litmus test. People with insurgent mindsets are in awe that Zuck would continue to take such bold risks, while the incumbent mindsets laugh and say “of course that would never work” and move on. Except the 10% of the time when it does work, and then the incumbents are eaten alive.
Ridicule aside, what’s the alternative? Facebook ultimately has a clear ceiling on their growth potential: Apple. Facebook do not own their own platform. You can dress it up however you want, but they are Zappos, and Apple is Amazon. Facebook needs some way to own their own platform, or they’ll always be subject to the latest ad tracking change from Apple. It’s why Meta poured billions into Oculus. It’s why they continue to develop AR smart glasses. They’re trying to create a world in which phones are not the future. They need to, or they die.
It’s easier to invent the future than to predict it
— Alan Kay
Engineers often think that if they just think about the system hard enough, very few iterations will be needed. For some types of companies, that’s true. Spaces where requirements have been very clearly mapped out, or if a very low level of innovation is necessary. This is usually how Big Tech operates. By the time you’re as big as FAANG, you know your customer extremely well, and you know what you’re building.
That’s patently untrue for startups. No software company in the world has Flows. We’re going to try random stuff, and see what sticks. Don’t treat iterations as a bug. It’s good to try some things that you think will be bad ideas because occasionally, you’ll be wrong. And more importantly, if you bet big, the reward for these will pay for all the mistakes you made along the way. By the way - all the pain we’ve taken is clearly paying off. Flows is the #1 reason, even more than auto scheduling, that businesses are signing up for Motion. It’s early, but it’s looking like we have another game changer on our hands.
I don’t mean to be cynical here. Critics are often well-intentioned. They believe they have some genuine insight, and by telling you “this won’t work,” they are saving you time so you can move on to other things that might work. What they’re missing, of course, is that nearly every idea has some critic somewhere. And the ideas that have no critics are typically useless.
Saying “this won’t work” never resonates with insurgents/founders because insurgents know that most things don’t work, and you can’t really know until you try. More importantly, the “this” that won’t work isn’t the end point. It’s the start of a series of iterations to discover what will work. Yes, users will hate it. But why will they hate it?
So to the founder, “this won’t work” isn’t some new insight from the heavens. It’s just one more piece of negativity they have to shrug off.
1 This isn’t strictly true. If the risk is so obvious that all of Motion would crash (a bad database migration, a self-DOS, a gaping security vulnerability) and cause an extended sev0, then obviously I have some veto power over that. But I would argue that 99% of the small disagreements I see in PRs aren’t about that. These are small exceptions far and few between.
2 For those who weren’t at the Chicago or Mexico offsites, the S curve is the hyper-growth typical curve that most businesses go through. It starts out extremely flat as you bash your head against the wall for a long time. Then, you find the key insight and exploit it to unlock rapid growth. Eventually, the TAM gets reached and you flatten out again. The key insight is that great businesses continuously find new ways to unlock new S curves. Our first S curve was with the consumer product, and we’re trying to unlock our second S curve with B2B now.
3 PMv3 was the first time we launched Pivot Table and completely re-did our PM pages. Hacker House was a 3 month onsite experience where a handful of us lived together in a house in Berkeley and worked 6 days a week, shipped super quickly, and iterated to find what ultimately became Flows aka Workflow Templates. We typically worked 80 hours a week during this time, with only Sundays being days off.
4 Separate from all this, do read more about Dmitry - he’s an incredible founder and amazing human being. An amazingly well balanced, well intentioned person who is humble and thoughtful while being ultra determined. Flo Health came into the market three years behind competition that was much more well funded (including Max Levchin from the PayPal mafia).