In Segment’s early days, we hit countless problems as a founding team. And at the time, I thought those problems were unique to our own special snowflake of a founding journey. I chalked it up to us being new grads and first-time founders.

But as I've worked with more and more startups, I've realized just how wrong I was.

Over the past five years, I've made about 25 different seed-stage investments. In doing so, it's taught me a LOT about the common errors that startup founders make. Even across different industries and levels of experience, I see founders hitting the exact same set of problems we encountered in the early days of Segment!

This post shares a handful of the top lessons that benefit today’s early stage founders. It’s a list of the things I wish we’d figured out earlier at Segment. [1]

1. Set goals

This sounds incredibly boring... but the #1 mistake I see startups making is that they don’t set goals. If you take one thing from this post, it's that you should set goals for where you want to be.

For the longest time at Segment, we didn't have goals. We moved ahead in various (often random) directions, and we would launch features consistently... but we never really set goals at all.

As we grew the company, we started to lose momentum. Teams were spending time on a bunch of stuff that frankly just didn't really matter to the overall business.

It wasn't until we hired our VP Eng, Tido, that we finally started setting focused and audacious product goals. Just by verbalizing where we wanted to go and then grading our results, our velocity improved by an order of magnitude.

I don't care if you call them OKRs, sprints, or something else entirely. Just set a deadline when you want to have something done and a metric you want to move or some other concrete result.

When early stage founders do attempt to set goals, I often see them agonize over what specific goals to choose. In practice, a "pretty good goal" is way better than "no goals at all". Perfect is the enemy of good.

If you don't yet have product-market fit, your goal should probably be getting your first 3-5 customers using the product. If you've hit the level where you now have dozens of users, your goal should be growing by an order of magnitude. It's better to get in the habit of setting and driving towards goals rather than being too worried about their exact semantics. Worst case, you just pick a better goal later.

A great set of goals answers the question: "what would have to be true in order for us to feel good about our progress at the end of the month?" [2]

If each teammate can independently answer "what are our goals for the month?" in the same way, you'll know you've succeeded. If you’re looking for prior art, read Measure What Matters.

If you strictly focus on making things true by a certain date, you are bound to make at least some progress towards your end result.

At the end of the day, every billion-dollar startup is really just the sum of many small deltas.

2. Send investor updates

Let's get one thing straight: your investors won't know everything. But investors won't know anything if you don't keep them updated on how things are going.

For the longest time, we were fearful of our Segment investors, to the point that we wouldn't bother sending them emails unless they asked about us. I can say now with confidence that this was 100% the wrong approach.

We worried that investors would think that we were screwing up (true) and failing (true as well!). And while that might be the case, a founder-friendly investor won't think that way. Investors, especially angels, invested because they believe in you. If I didn't think a team would go somewhere, I wouldn't put money in.

With each investor, include the goals you're working towards, as well as the asks for them. I think monthly is about the right cadence for this in the early days, moving to quarterly around Series A/B time when you start partnering more with a few board members.

Simply writing the updates should clarify your own thinking tremendously. If it takes you more than 1-2h to put together an update on the most important things happening at the company, it's probably a sign that you should be doing more thinking about the big picture.

When asking for help, the things that our investors have been able to help us with evolved quite a bit over time. But here's a good rule of thumb for what to ask for:

  • Seed/Pre-seed: user-testing, intros to beta users, hiring

  • Series A: hiring, early customers, go-to-market

  • Series B/C+: comparables at other companies, senior/exec hires, specific expertise around things like management, infrastructure, systems, etc.

Not all investors will be able to help with everything. But at the very least, sending them information will help you be top of mind. I've lost track of how many times a moment of serendipity where I'm catching up with an old friend has led to a meaningful conversation for a company I've invested in.

As an extra bonus, the strongest startups send these updates to everyone on their team. It's amazing how much putting the goals in writing helps everyone stay on the same page about what's most important.

One other note here: take pictures. I now wish we had far more pictures of Segment at every stage of the company. They help turn investor updates into cherished memories.

3. Launch (bonus: self-service)

Okay, this lesson is taken directly from the YC playbook. And YET, I see so many founders (including YC founders) fail to launch their product.

If no one notices your launch... just ignore it and then launch again. If you're doing things right, you'll never run out of stuff to launch! [3]

Why is launching so important? Let me share a personal story...

We spent about 1.5 years building different iterations of analytics tools. For every iteration, we had a waitlist that users could sign up to use. We personally reached out to the users who we thought were the best fits, and then tried to set up time to use the product.

The result? Nobody cared. We had no users. We never launched.

When we finally launched Segment in it's current incarnation, we threw out that approach entirely. We put up a self-service flow, and let anyone who wanted to sign up for it.

That's when something strange happened... we attracted an entirely new set of developers who just were crawling out of the woodwork and excited to use the new product we'd built. They were coming from companies far outside of SV that we'd never heard of.

It floored me.

Lesson learned: the people you happen to be talking to now are probably not the people who have the biggest problem in your space. Do everything you can to reach the folks with the biggest problem, and then, reduce any barriers they might encounter.

I expect a bunch of you reading this post to ignore this advice, just like we did in the early days. It takes a certain confidence to launch something you've built and put it out there for the world to see. Ultimately though, the rewards are worth it. You'll see users coming from communities you've never even heard of.

4. Consider working in public

Let me tell you a tale of two startups.

Startup A is constantly putting up interesting content on their blog. Their founders are sharing product launches, engineering posts, and creatively brainstorming about what it takes to solve problems in their market.

Startup B is operating in stealth. You can't find much about them online, but one of the founders reached out with a nice personalized email mentioning their funding by a top-tier VC.

Suppose you're looking for a job... do you pick Startup A or Startup B? In my experience, A almost always wins. Momentum is a compounding force.

Unless you are working in a space that heavily depends on IP, you should probably be publishing more content about what you are doing. This could be open source, it could be a weekly newsletter, it could be a changelog. [4]

Whatever it is, it's going to help you both hire and attract customers. So much of the internet is merely about consuming, that just by putting ideas out there, you’ll have a leg up on the competition.

In the early days of Segment, our user acquisition was powered by open source projects and blog posts. But I've seen founders have success with Twitter, Substacks, Podcasts and a variety of different channels.

If you're looking for inspiration, the Railway and Linear changelogs are epic examples. Companies like Baremetrics and Buffer differentiated themselves by just being open and honest. Stripe and Figma's blogs not only share what they build, but how they built it.

5. Know when to hire

There are three big inputs that all startups need to continue growing

  1. product capabilities

  2. customers and users

  3. hiring

I've put them in roughly the order that most teams encounter them.

Startups begin with a small product that has a modicum of utility. It starts attracting a handful of customers organically. And as more and more customers start using the product, the founders realize that they need extra help to stay on top of all of those requests.

Remember though, the real goal here is #2. It’s not the size of your team, it’s the value you’re able to provide to the world. [5]

I see hiring a big team before you have product-market fit as a red flag. If the company doesn't yet have a set direction, it's going to be harder to pivot with 10 people than it is with 3.

But, I've worked with startups who have traction and runway... and just seem to be spinning their wheels under load from existing customers.

I get it. Hiring isn't the most fun, especially for an introvert. It's a lot of interviewing and feeling like there's a lot of rejection. Rejecting people sucks. Losing candidates sucks. For many of our roles at Segment, we've had to talk with 50-100 different people to make an eventual hire.

If you have 24 months of runway, and a clear list of things you’d do if you had more copies of yourself: you probably aren't spending time to hire the people you need.

At Segment, my co-founder Ilya fulfilled this role. It was back in 2014, we were 12 people at the time, had raised a $10m Series A, and had $1m in revenue.

Things were good... except for the fact that we were getting overwhelmed by the number of support tickets coming at us.

At one of our weekly founders dinners, I remember Ilya making the case that we were really at a pivotal moment. "If we don't start hiring people now, we'll hobble along forever as a 10-15 person company. We have to make a change."

Ilya was right. He started a six-month focus on recruiting. He built our initial success engineering team, our sales engineering team, and our CSM team. His work allowed us to hit 2.5m in revenue in the next year.

There are a few other founders I talk with who are in this same situation. They raised money, have a lot of runway, are overwhelmed. For those founders, I say be the Ilya you want to see in the world. If hiring is a problem go after it with tenacious force.

6. Optimize for learning

I model most startups as a continuous voyage of discovery. It's one of the things that makes creating a company really rewarding; the problems are often novel and exciting.

The startups who most rapidly improve are the ones who are able to take outside information, and then incorporate it into their product, their strategy, and their worldview. That's why, you should always be optimizing for one metric in addition to revenue and user growth: learning.

At any given time, your startup should have somewhere between 1-3 existential risks. Big questions that you are worried about that you really want the answers to. It's your job to get the answers to these questions.

To do that, I've seen a couple of tactics that really work. The first is running a pre-mortem. Imagine yourself in 18-months, you've run out of money, and it's time to shut down the company. What made you fail?

Once you have your best guesses at an answer, tackle those problems first.

A second very useful tool: ask each person to write down individually what they think the biggest risks will be, and then share them. I've found time and time again that each person will bring a very different perspective to a big project. What one person doesn't see, another will highlight. It's systematically de-risked blind spots on numerous teams I've been on.

7. Pick the partner, not the firm

When you are ready to go raise your Series A (or even your seed stage funding), weigh the individual partner you'll be working with far more than the reputation of the firm.

Sure, it was useful that we raised our series A from Accel, who had a great reputation and a large portfolio we could cross-sell into. But what I wouldn't trade is the caliber of advice we got from each of our investors who joined our board: Vas, Will, Miles, Ali, and Rob.

At each step of our fundraise, we optimized for the partner we would ultimately be working with far more than how much of a name-brand the firm is. If you are building a startup for the long-haul, be aware that it really is a 10-year+ journey.

The people you have in your board meetings matter far more than the email domain they use.

8. Focus

I see a lot of companies out there trying to do too much. They often are trying to accomplish A, and B, and C at once... without really asking themselves "what's most important: A, B, or C?"

As a startup, the ability to focus is your best advantage against a big incumbent. Any one of your bigger competitors will have many more resources (both money and people) than you do. Your competitive edge is the ability to tackle a narrower market or use-case, and grow your business from that narrow wedge.

For most startups, it's better to do a few things well than try and do too many things poorly.

With Segment, we eventually found focus by just launching a hundred-line javascript file. The use case was so narrow, it almost seemed like a toy. But focusing on just that one point of collection gave us the momentum to steadily expand to other areas of collection.

We didn't always nail this. At times, our focus diverged, and we found ourselves a bit overextended in the use cases we could support. We didn't have a clear understanding of exactly what it was that we did for customers, and what use cases we should really double down on.

But when we did focus as a company on some singular goal, be it revenue, churn reduction, or a big product launch—the goal usually happened.

The startups who do this best apply an almost maniacal focus to solve a narrow problem incredibly well. The ones who dominate always can answer the question: “what’s most important?”

The view from the other side

I’d like to leave you with a final parting revelation I’ve had as an angel investor.

While founding a company, I misunderstood the goals and motives of most of our early angel investors. I thought it was purely about financials.

Now that I'm on the other side, I realize that the best thing about being an angel investor is that you can have a significant impact on a startup's trajectory. It's not anywhere near the impact that the founders will have... but sharing experience and advice can really make the difference during pivotal moments.

We didn’t leverage this expertise nearly enough in the early days of Segment… but when we did, the advice shared by the YC team (PG, Sama, Harj) and our angel investors was invaluable.

In retrospect, I think the most valuable thing our early investors did was bring a perspective that was "just far away enough" from the problems we were facing.

We'd tell our investors all of our problems... and they'd say: "why don't you just do X?". And we'd realize: "oh yeah... why don't we do X?!" They helped us stop overthinking, and start taking the obvious path forward.

I want to repay that favor to founders I work with, by being a sounding board, a gut check, and an outside perspective. I don't want to micromanage how you should run your business (that's your job). I don't want to ask for my money back. I want to help.

Most angels I talk with, especially former founders and operators, seem to feel the same way. We’re excited to help the next generation of founders drive innovation, one step at a time.


[1]: Over the past decade, there has been a wealth of great lessons shared within the startup community. For further reading, see: The Hardest Lessons for Startups to Learn and How to Start a Startup.

[2]: As an aside, I'm quite biased, but I believe Segment is the best way for most early-stage startups to track their goals and usage. Beyond that, setting up a spreadsheet that you update 1x/week or a SQL dashboard with PopSQL/Mode/Metabase all work well too.

[3]: If you are a hardware company, launching early gets harder, but the wisdom is still the same: launch a first (small iteration), then launch a bigger iteration, then launch a bigger iteration.

[4]: If you're looking for one of these pre-generated changelogs, a friend is creating makelog.io

[5]: This doesn't have to be the goal necessarily, some startups are more art or passion project than anything else. But, I find for most startups this is approximately what the end-game is, especially venture-backed startups.


Thank you to Kevin Niparko, Victor Pontis, Danqing Liu, Jake Cooper, Evan Conrad, Tejas Manohar, Kevin Liu, Tido Carriero, and Peter Reinhardt for providing feedback on drafts of this post.