If There Aren't Any Typos In This Essay, We Launched Too Late! If There Aren't Any Typos In This Essay, We Launched Too Late! If There Aren't Any Typos In This Essay, We Launched Too Late!

Reid Hoffman

In my early years as an entrepreneur, the world was still largely ruled by scarcity. Products required raw materials to produce. Distribution was expensive and logistically complex to achieve. That led to a mindset of prudence and precision expressed in aphorisms that still carry a great deal of weight today.

You only get one chance to make a first impression.
Measure twice, cut once.

In a world ruled by gatekeepers and beholden to physics, this attitude made sense. Space on retail shelves was finite. Creating the machines and processes it took to produce products at a scale that made them affordable to mass markets was expensive and time-consuming. Through the mid-1990s, even with something as virtual as software code, this was true. In those days, software was not so much eating the world as adding to it. Shipping Microsoft Word or CorelDraw required even more atoms than shipping a novel or a record album. Software products came on disks, in boxes, with thick user manuals. They were costly to store, ship, and merchandise.

In this world, where getting the latest version of your favorite spreadsheet program necessitated a trip to the mall and a significant cash outlay, it made sense to put significant effort into getting things right from the start. If you didn't, users could rightly insist that you harmed them. You cost them time. You cost them money. The loss of trust that a less-than-ideal interaction created could last for years. Or maybe even forever.

Now, everything's different. Software doesn't come in boxes anymore. Laptops don't even have disk drives.

Whether it's through search, social sharing, or virality, it is now much easier and less expensive to reach potential users. Users, in turn, no longer have to pay hundreds of dollars to try a new app or service. This new world marginalizes the cost of second impressions, third impressions, and even tenth impressions. As a result, the value of getting it right the first time has plummeted. And a new batch of aphorisms has emerged to crystallize the new ethos.

Move fast and break things.
Fail fast.
If you're not embarrassed by the first version of your product, you've launched too late.

The latter is one I coined myself, more than a decade ago.

While the point of an aphorism is to convey a rule or advice in a crisp, standalone way, I continue to get questions from people about it. Some challenge its assumptions. Others simply want more elaboration.

There were three distinct themes informing this statement. The first and most explicit one is the importance of speed: You should aim to launch fast.

The second, less apparent theme, is that your assumptions about your customers, and how they'll use your product, won't be entirely correct. Eventually, you'll probably be embarrassed by how many wrong assumptions you made.

The third theme is that you lose out by delaying the onset of the customer feedback loop: If you'd launched sooner, you would have started learning sooner. Instead, you launched too late. The word "embarrassed" plays a key role here.

Some people may interpret my aphorism's emphasis on speed as permission to cut corners, act recklessly, or proceed without a clear plan. But I didn't write "If you're not indicted by the first version of your product, you've launched too late."

Nor did I write "If you're not deeply ashamed by the first version of your product…."

Indeed, if you launched so fast that your product generates lawsuits, alienates users, or burns through capital or other resources without any apparent gain, you did in fact launch too soon! But what if all that happens is that the user says, "Well, that was lame…."

Sure, you might be a little embarrassed, but that's not a bad thing, as long as you're moving fast, learning fast, and improving your product.

Even in our new world, many entrepreneurs still identify so closely with their products or services they cannot bring themselves to adopt a mindset that prioritizes iteration and learning over perfection. As an entrepreneur, you often base your whole sense of identity and self-worth on your "baby" – and that almost guarantees a sense of driven perfectionism. You believe that you are going to be judged on your product – or even equated with it – so you want everything to be exactly right upon the initial unveiling.

In reality, you should be moving faster than that.

If you're willing to be embarrassed, you gain speed and flexibility. You get to market faster when you're not trying to anticipate and design every feature and potential use case. That means you also start collecting feedback from users faster, and learn about what happens in instances you didn't or couldn’t anticipate.

In the end, entrepreneurs shouldn't be so focused on what happens on the first day of launching a product. The relevant timeframe you should be thinking about is the first month, then the first year.

Also, as Eric Ries once suggested to me, there's "a secret corollary" to my aphorism. Namely, that "no matter how long you wait to release your first version, you will be embarrassed by it." In other words, even if you blow deadlines and exceed budgets in an obsessive effort to achieve perfection at launch, you're still going to be embarrassed by your first version anyway! It's inevitable, because it's impossible to exactly foresee whose going to use your product, toward what ends, under what conditions.

For example, in the early days of PayPal, we thought that allowing users to "beam" money to each other using their Palm Pilots – so they could split a lunch check, say – would be our defining feature. In fact, we ended up launching late because we put so much energy into building this functionality. As it turned out, though, what users really wanted was a way to exchange money using email, to facilitate their eBay transactions.

In the case of LinkedIn, our founding team had a lot of debates about whether we could launch without a feature we'd dubbed "Contact Finder."

"Contact Finder" was conceived as a messaging service that would propagate through your contacts list. You'd send out a message saying you were looking to hire a front-end developer or a marketing writer, and your contact would forward the message to their contacts. Eventually, a front-end developer or marketing writer would see it and then connect with you. In early 2003, there was great concern amongst our team that without this feature, people wouldn't truly know how to use LinkedIn. But there were other factors in play that made it desirable to launch sooner rather than later.

We had to think very hard about what constituted an acceptable first experience with the site. We knew users would be able to send invitations to people they knew. We knew they'd be able to do basic search and communicate with others.

"Contact Finder" didn't exist yet, but ultimately we decided it was more important to launch sooner than it was to have that specific functionality from the first day. And we were right. In fact, 14 years later, we still haven't introduced "Contact Finder"!

By launching without it, we were able to start a dialogue with our users that much faster, and start learning from them. And while our team might have felt more confident had we launched with every bell and whistle we'd envisioned, the truth was that what we did launch with was enough. Our feature set was incomplete, but it wasn't so barebones or so flawed that it created a negative impression. Instead, there was enough there to get users to engage with the platform, and that's all we needed.

Once they were engaged, we started getting feedback and iterating.

That is the ultimate goal of my aphorism: To encourage a culture of flexible persistence and iterative improvement. For that, a little embarrassment in those first few weeks after lunch is a small price to pay.


Note: Anyone is welcome to republish this essay.