Things that aren't progress

A few months ago, I wrote about things that look like work, but aren't. As I paid more attention to founders doing these things, I started thinking about why they were happening. I realized that the behaviors were largely a function of bad goal setting.

When founders choose bad goals, they create bad metrics around them and try to hit those metrics. This means that those founders are optimizing for things that look like progress, but aren't. This is dangerous for startups because founders that aim for bad goals warp everything they do towards measures of progress that don't help the business grow and succeed.

If you find yourself thinking that these things are progress, you need to examine what you're aiming for.

Things that look like progress but aren't:

  • Conversations with large enterprises - This is a big one for companies that rely on enterprise sales. Usually, these "conversations" are non-binding, low level, and scattered. They rarely lead anywhere. There is a type of conversation that is actually progress, but it is generally paired with a clear sales plan, a list of stakeholders, and clear asks met along the way.
  • Press mentions - This feels so good, which is why it is tricky. External validation is awesome! But it isn't as awesome as engaged users or paying customers. If you're looking to get press as a means of building SEO, then press mentions can constitute progress. That only makes sense if SEO is part of your strategy and you're measuring things properly. Russ D'Souza talked abut how they do this so well at Seatgeek:
  • Winning awards - This is another one of those pieces of external validation. Certainly helpful when proving to your mother that you're not wasting your time, but not material to the success of your company. Awards may recognize either the progress you've made, or how good you are at PR. Focus on making progress.
  • Getting retweeted by someone famous - This is similar to press mentions, but even worse. When a founder is trying to do this, they're usually working on their personal brand at the expense of their startup. That's definitely the wrong thing to do.
  • Meeting someone famous - Famous people are really interested in startups these days. It feels good/cool to say you met someone famous. Don't underestimate the value of having great experiences, but don't mistake it for a sign of success.
  • "Getting into" elite conferences, like Davos - This is some strange combination of awards and meeting famous people.
  • Eyeballs - Had to include this because it's part of what drove valuations in the tech bubble. Turned out this wasn't progress. There's a corollary today: uniques. That can be a good measure of progress if you're a media site generating revenue off of impressions, but in almost all other cases, it's a vanity metric.
  • Cumulative registrations - The cousin of eyeballs. People who register for your company but don't use it aren't actually helpful. In fact, this is probably a bad sign because it means that people don't actually want to use your service. (Suggested by Geoff Ralston)
  • Hiring - When you hire for the sake of a world class team, that isn't progress. If you're hiring because you can't handle the load on your service or product, then it probably is a good sign and it is progress. (Suggested by Paul Buchheit)
  • Fundraising - So many founders confuse raising money with success. That's how the press views it, but it's one of the worst things that founders can tell themselves. Fundraising is just a tool to accomplish your real goals, not a goal in and of itself.
  • Getting into an accelerator, even YC - This is a corollary of fundraising and awards. At YC, we'll do our best to help you figure out what your goals are (if you don't already know), and work with you to achieve them, but it's what happens after you get in that counts, not getting in. YC is neither necessary nor sufficient for success.

I and We

Starting a company is, in large part, an act of ego. When that business is a startup, the ego component is even larger. This is a good thing.

Ego is what gives the founder the confidence to create something new. Ego powers the belief that the new thing the founder creates will be good enough to change the way that tens, thousands, even millions of people live their lives. Maybe we don’t always call this ego -- maybe we call it vision, or confidence, or passion -- but the idea is the same.

But as important as ego is to the founding of a company, it is also corrosive to the creation of a good culture. Unbridled ego becomes arrogance. It doesn’t allow for other people to achieve and contribute. Founders who do not keep their egos in check are unwilling to acknowledge the help given to them by others, and have a hard time building and retaining great teams.

To start a company founders need ego, but to build a great company, founders need to be modest. Modesty is what allows founders to see all the things that contribute to their success, especially the things over which they have no control. With modesty, founders can see the important role of luck in their success. They can recognize, acknowledge, and reward the part played by cofounders, employees, and customers. They’re resilient when things go wrong because they can see beyond themselves.

The interplay between ego and modesty is obvious when you hear founders talk about their companies. As a company grows, the best founders increasingly talk about “We” and not just “I.” Every discussion about the achievements of the company is a chance to highlight the contributions of other people and of the organization overall.

This isn’t to say that the best founders disappear into the background of their own companies. The “I” still plays a strong part. It remains the founder’s job to consistently set the vision and the example for every employee. When that goes away, a company risks moving into stasis or decline.

Perhaps the most important place to use “I” over the life of a company is when things go wrong. When figuring out what happened, and who bears responsibility for fixing a problem, “We” is insufficient and even dangerous. In that case, modesty doesn’t drive the use of “We,” arrogance does. Arrogance won’t allow a founder to admit that he was wrong, so he slides to “We” to cover for it, to blame the organization. But diffuse blame means that no one figures out what actually happened and how to fix it. Do that often enough, and the badness grows until it kills the company.

It’s tough to balance “I” and “We” and it gets harder as a company grows and becomes more successful. As a company does better, the easiest stories for the press to write are those about the genius of the founders. At the same time, as more and more things happen at a company that the founders don’t directly touch, founders may start to feel disconnected from the thing that they created. This is a hard thing to face, and some founders respond to it by claiming sole credit for successes that are the work of many people. From the other side, as the company grows, some founders fade into the background and stop providing the singular vision and leadership that the company needs to succeed.

There doesn’t seem to be an easy answer as to how to strike the right balance, nor is there a single paradigm for what works best. What does seem clear, though, is that founders need to keep these questions in mind, and, if they find themselves only using “I” or only using “We,” to think long and hard about what they might be missing.


Thank you, Colleen Taylor, for your edits.

Presumption of stupidity

I've noticed a common bias that shows up in some founders: they believe that their competitors are stupid or uncreative. They'll look at other businesses and identify inefficiencies or bad systems, and decide that those conditions exist because of dumb decisions on the part of founders or employees. 

This is a bad belief to hold. In truth, competitors in the market are usually founded and run by intelligent people making smart and logical decisions. That doesn't mean that all the decisions they make are necessarily the right ones, but they're rarely a function of outright stupidity.

Where companies do things that diverge from what seems smart from the outside, it's a much better idea to ask why those companies are doing things from the presumption of intelligence and logic rather than the presumption of stupidity. If you don't ask these questions, you might find yourself making the same decisions, or ending up in the same place with your own set of rationalizations. I see this all the time.

In fact, we made this mistake when we started Tutorspree. We looked at all the local agencies and the way that they acquired customers and charged for packages of lessons. We assumed they asked for so much money up front because they were greedy and not smart enough to figure out a better system. It turned out that packages of lessons were a logical outgrowth of high upfront acquisition costs and the long term dynamic of tutor/student relationships. A large enough subset of customers appreciated the breaks on pricing and commitment created by booking multiple lessons up front that it made sense to model the business that way. It took longer than it should have to realize this because of our bias.

If, instead, you presume intelligence and analyze the reasons a business looks the way it does, you will often see the challenges you might face ahead of time and, as a result, design a solution that is actually better, as opposed to simply looking new. It is a lot harder to think this way because it means that you can't just dismiss the things other people do and assume you'll be better. You actually have to prove that you know how to be better. That can be really scary because, much of the time, you might not be able to figure out how to be better. Everything you think of might lead you to the same place you see your competitors.

That, though, is no reason to stop working on your company. I think it's actually a reason to keep going, and to keep gathering information and generating new ideas. This is part of what's so cool about starting a company, you get to make up new rules as you go along, and you can toss out old ones as you go along. Two founders looking at the same problem can easily come up with multiple solutions. Each solution might look similar from far away, but the small differences add up. Importantly, if you know that other smart people started in a similar place and ended up with the wrong answer, you'll think a lot more critically about each of your decisions and never get lazy about challenging your own assumptions.

Of course, just because you presume intelligence doesn't mean that every decision made was smart. People and organizations make bad decisions for all kinds of reasons. The thing is, you don't learn much by understanding that a call was bad, you learn by understanding the inputs and the organization that enabled the bad decisions.

Even with this framework, there's no guarantee that you'll end up in the right place, no matter how much you analyze those whose decisions have left you with an opportunity. At the end of the day, there's only so much you can learn from looking at competitors. Truly great businesses aren't built as counterpoints to existing companies. They become great because they meet a deep need that isn't being satisfied. That usually requires the kinds of creative and cognitive leaps that no amount of market analysis could possibly give you.

Private infrastructure

When infrastructure is built, it usually starts out as a large scale project that can only be accomplished by government. It can be built in undeveloped areas for a fairly large amount of money, or in developed areas with massive amounts of cash and even more political capital. That's hard to do in democracies, though seems to work well in places like China.

This dynamic means that we'd expect infrastructure to fail over time as the inertia arrayed against repairs and new construction grows. That seems to be what's happening in the US.

When infrastructure decays and fails, though, it creates a lot of opportunities. Citizens who were meant to be served by the infrastructure become unhappy and are when that unhappiness is great enough, they will spend their own dollars on alternatives. This is a more direct process  than funding infrastructure through tax dollars, though it can produce different types of outcomes.

In the United States today, these opportunities are being exploited by a range of startups and companies who are essentially building private infrastructure, mainly for the relatively wealthy.


Instacart is building new infrastructure to deliver food to homes. In most early use cases, this looks like pure convenience, however, over the long run it could impact and start to eliminate food deserts. By making food available in more places, Instacart has the power to change where people want to live. That process will reshape how cities grow and, eventually, are built. I'll also say that in a place like NY, where supermarkets are small, crowded, and overpriced, it seems clear that the current system only exists because there's no better alternative. Instacart is that alternative.


This process is also starting to play out on our roads. Most of our highway system was created by a single gigantic bill, the Federal Aid Highway Act of 1956. It would be hard to imagine today’s Federal government getting its act together or finding enough money to repair roads on a national scale. Harder still would be believing that a place like the Bay Area would figure out a way to install effective and efficient public transportation to ease the stress on our roads and highways which are nearly always gridlocked. Technology is starting to provide ways to bypass those problems by improving throughput on existing systems without changing the physical plant.[1]

Self driving cars are the most extreme example of this trend. If we do actually arrive in a world where self driving cars are ubiquitous and built on the same standard such that they can communicate with one another at long range, congestion gridlock should largely vanish as the cars plan miles ahead and subtly change speeds to clear up slow downs before they start. Even before that future, though, apps like Waze help drivers plan better routes and alleviate some stress on the most crowded points in the system.


I think the trend is also impacting power generation and consumption in the US. Companies such as  Solar City will install your own power plant on your roof which means you're no longer subject to brownouts or price increases at peak times. In most of the US, this is more of a ‘nice to have’ and cost saving measure, but in the third world, solar power could provide a viable and efficient alternative to the construction of large power plants and electric delivery systems. These systems would be a fraction of the price of the old way of doing things and be significantly more reliable thanks to their distributed architecture.[2]

Private can’t mean wealthy

There is, however, an underlying tension in the rise of this new infrastructure in that the first segments of the population to get it are generally the wealthier ones who can afford to use it. Two tiered systems may be fine when they exist in non-essential parts of our lives. However, when there are two tiers, determined by wealth, for services as basic as energy, we're in a dangerous place.

That's part of why infrastructure exists the way it does. No single piece of the population would spend enough money to build services and systems useful for everyone. I think that's always going to be true for certain projects, such as bridges. But, as technology drops the cost of delivering the types of services we traditionally associate with infrastructure, I think we'll see the market extend its reach farther and farther. There's a huge amount of money to be made by ensuring universal coverage, but it won't be easy to unlock all of it.


[1] At best, though, this is a partial solution. Throughput is well and good, but if the roads actually start to collapse, we'll need other solutions.

[2] Even though this seems like a good idea, there are still quite a few hurdles, not the least of which would be figuring out the right financing structure to make the installations profitable for the companies doing them. This is harder than it might seem for developing countries without sufficiently developed/ubiquitous banking systems.

We need to rethink employee compensation

I think that the way that employees are getting compensated at startups is starting to break. The old model of relatively low salary and "high" equity only works when there's a healthy public market for that equity in the not too distant future. If IPOs are getting significantly delayed and are potentially at risk of not happening at all,[1] we need to change how compensation is structured.

Options only make sense as compensation when you can draw a clear line from the point at which you get the options to the point at which those options are equivalent to cash. This happens when two conditions are met: 1) the stock price is enough above the strike price on the options to make exercising worthwhile to the employee and 2) there is someone willing to quickly buy the stock. This tends to happen with access to liquid markets, which traditionally means public ones. That's the only place where an employee can sell whenever she wants and can know the price she'll get with a high degree of certainty. It's also the only place where an employee can really hedge out some or all of the risk of her options.[2] That hedging is important if the value of the options rise significantly and lead to an employee having 99% of her net worth in a single illiquid asset.

When shares can only be sold in private transactions on secondary markets, and, increasingly can only be sold with the consent of the company, the options are actually worth less. This is true for three reasons.

The first reason is that control is an important factor in asset pricing. Think about this in a situation where an employee would like to sell stock to fund the purchase of a home. In a public company, that employee could exercise the options, sell the stock quickly and use the cash to buy that home. If that employee instead wants to sell stock in a private company, and the company does not let the employee do so, the employee cannot get the cash to buy the house. Because that employee cannot do what they want with the equity, that employee should discount the value of the options they hold.

The second reason that options are worth less in the absence of robust public markets is that the terms of sale of stock in the future are unclear. An employee being granted options has no idea under what conditions he will be allowed to sell stock in the future. The future sale might be subject to board approval, there might be a heavily restricted set of buyers, the employee could be told that no sales are allowed whatsoever for a variety of reasons. Each of these possibilities adds uncertainty which makes the options harder to value and worth less.

Founders need to account for this change in the value of options when giving them, and employees need to account for the change when evaluating their offers and negotiating. This is, however, hard to do in the absence of information. Without a liquid public market to value stock, it's hard to say what the options are worth. The best guess an employee has in this scenario is to go by the most recent valuation given to the company during a fundraise. The further from that fundraise an employee is, the less relevant the price. Furthermore, without the ability to compare how different employers control and handle the sale of private stock, it's nearly impossible for an employee to know what is standard.

The third reason for why individual options are probably worth less now than they used to be is that both employer and employee need to account for the fact that the time until IPO or liquidity is longer than it used to be. This is a big issue. To get the true value of offered comp, employees need to add their offered salary to the present value of the options offered. When calculating that, the further out the payout, the less it is worth today.[3]

This is an especially important calculation when trying to hire employees away from larger companies who are already public or can offer larger cash packages to employees. Those offers also tend to include options or stock grants which have known price and sale terms. Because of their easy to understand high dollar value, these offers are hard to compete with. Options in early stage startups are much harder to value even without all these new factors. Given uncertainty around the final value of those options, the time until they can be exercised and sold is hugely important. As you move further out on the curve, not only does the discount factor become a larger lever, uncertainty over the final value gets more extreme. You can be pretty sure that a company currently worth $10mm won't be worth $1b in 3 months, so you have a reasonable band of expectation. Once that value calculation moves to years out, nearly anything is possible and needs to be factored into the calculation.

I'm not the only one thinking about this, and, indeed, there are instruments available that let employees take immediate advantage of long term options. Some banks or private investors will provide loans collateralized by options. This can be good, but debt is very different than having cash. Taking advantage of this kind of strategy could also lead to potentially tricky situations where, if the option value as determined by secondary transactions falls (which could happen quickly and dramatically given the lack of liquidity in the system), the creditor could claim the assets before the term is up. Even if the price then recovers, the employee is out of luck and money.

Founders need to deal with these issues, and given how fierce the competition for good employees is, I think someone will do so soon. One response would be to increase overall options packages to account for the fact that they are now less valuable.

Another path would be to create an internal liquidity pool capable of buying shares from employees at predetermined trigger points. Maybe this could be funded by existing investors, or maybe by other employees or existing shareholders. Because the sale points are predetermined and don't provide an open market, the value of options would still be discounted, but not as much as a scenario with no guaranteed access to liquidity.

It may make sense for some companies to start creating revenue share programs as standard practice - especially for companies that never plan to going public (of which there appears to be a rising number). This could be a powerful means of retention, though would only work with companies throwing off enough cash.

I think that something even more basic needs to happen first: companies need to standardize their secondary sale and options repurchase practices. We're living in the wild west right now - each company and each employee is doing things on their own. An employee that wants to sell shares on the secondary market generally starts asking friends if they know anyone who wants to buy those options. Generally, they'll find a broker who offers to matchmake and who then gets a cut of the overall sale in exchange for their work. The price that gets set isn't shared from one deal to another, and the rules around whether or not the company can block the sale aren't standardized. This isn't fair to employees, and will lead to confusion and bad outcomes.

Maybe what we need is two things. The first is a founder pledge that they will do everything in their power to let common holders sell into secondary markets above a certain valuation, say, $500mm. The second is that founders of highly valued private companies need to participate in a robust, standardized, secondary market to allow for the clearance of those shares.

My guess, though, is that there will be a number of different solutions with the market choosing the best one as evidenced by the quality of talent hired and retained by the company using them. Founders should be watching for that and be willing to adopt best practices as soon as they arise.


[1] There have been a lot of articles and blog posts about this recently, but Andreessen Horowitz's recent "US Tech Funding - What's Going On" is the best analysis I've seen. They note that the average time to IPO is now 11 years vs. 4 in 99, and that the overall number of tech IPOs is plummeting as privately funded companies raise huge late stage private rounds instead.

[2] You can do this a number of different ways and hedge out either your exposure to the company itself or do the overall market. Hedging out this risk probably deserves an entire post as it can get complicated. And, while nearly impossible to do it perfectly, is important to significantly reduce risk.

[3] We'd do this as a lump sum calculation for simplicity. You can play with that calculation here:

Party (round) all the time?

There are a lot of party rounds happening right now in early stage investing. Definitions vary, but calling a round with greater than 10 investors a party seems about right. My thought is that party rounds tend to leave companies without an investor who cares enough or has pockets deep enough to bridge the company when necessary. Some rounds do actually have a strong lead, along with a syndicate of smaller, useful investors. Those generally have different dynamics.

What I've been trying to figure out is whether or not party rounds are increasing in frequency or not. Anecdotally, it feels like they are. Has something changed in the valley that party rounds should become preferred? The JOBS Act made it easier for large numbers of small investors to back companies in a way that was not previously legal. Platforms like Angel List reduce the friction required to find investors by creating a central, easily accessibly clearing house for companies looking for money and investors looking to give it to them. Or maybe the growing size and frequency of Demo Day type events (YC's included) have created an environment where party rounds are the new normal.

I took a look at Crunchbase, expecting to see that the overall number of party rounds has been rising uncontrollably. Interestingly, that's not what I found.

What's actually happening is that there are more startups getting funded, and the number of party rounds going to those companies is rising roughly in tandem. What's really surprising is that party rounds as a % of overall rounds actually fell from a peak of 3.7% in 2010 to 2.2% in 2012, though it's now back up to 3.42%.

So it seems that we're not seeing a new normal when it comes to early stage investing. We're actually seeing an overall stable pattern ticking up in the last year after a drop. The real story here is the one about the overall number of startups getting funded as the economy has come out of the Financial Crisis. That number is clearly rising, though 2014 tailed off a bit[1]. More startups mean more innovation, and that's great.

The structure and composition of a single round of funding isn't the most important factor in a startup's success or failure. It's just another piece of the story, and one that doesn't seem to be changing much.[2]


[1] Not sure why the number would have dropped in 2014. Might be a function of the data available to Crunchbase when this run of the db was pulled on April 3, 2015.

[2] I am still curious about whether or not party rounds are a positive or negative influence, but don't yet have an answer.

Title inspired by: 

Things that aren't work

Recently, I've had a few conversations with founders who, after YC is over, feel a bit lost about what to do next. During YC, the answer of what to do is pretty simple - we tell them to write code and talk to users. Intellectually, they know the answer after YC is to keep building their businesses, but all of a sudden they're faced with a lot of different opportunities, needs, and demands on their time.

This is a tricky time for founders because it can be easy to confuse things that look like work with actual work. Work may mean more than just writing code and talking to users, but it should only encompass things that make your company grow and get better. It can be easy to convince yourself that doing these other types of things make your company better, but that's wrong. These things are mostly for recreation. Treat them that way if you want to, but don't confuse them with what will actually help your startup.

Things that look like work but aren't:

  • Writing blog posts about running startups - This feels good. If it gets onto Hacker News and gets a lot of views, you'll feel really flattered and proud. But don't confuse people reading your post with people knowing and caring about your company.
  • Speaking on panels at startup conferences - If your customers are other startups, and the panel is about something specific you do, great. If not, this is a waste of time.
  • Going to fancy conferences hosted by investors or media - Feels great because  you get to talk to people about how well you're doing. If you were doing that well, you'd be in your office or talking to customers. Simple test for conferences: Are your customers or users there? If yes, could be worth going, if not, then it isn't.
  • Advising other founders - If you know enough to genuinely help, this is a really nice thing to do. It's good to help other founders, but it isn't likely to help your company grow. Make sure you are careful about this and don't let it take up too much time.
  • Investing in other startups - Definitely not work, though it might make (or more likely lose) you some money. This is on personal time.
  • Being a venture partner for a VC - Opposite of work. This is probably taking time away from your startup because now you're working for someone else.
  • Attempting to fix the plumbing in your office - I've done that. Not work. Not the best way to save money. My former cofounders would probably tell me not to insulate the windows myself either.
  • Networking happy hours hosted by investors - This is an opportunity to drink free beer. Great to do in moderation.
  • Having coffee with investors - This can be confusing, because sometimes you need to meet with investors. If you're gearing up to raise money or need specific advice, this is work. Most of the time, though, this isn't work.

Things that don't look like work but are:

  • Writing updates for your investors and meeting with them one on one - I've written about investor updates. These relationships are important, and can be incredibly helpful as you grow. Maintain them.
  • Talking to your cofounders and team - Sometimes, this looks like having coffee or grabbing a beer. Invariably, you'll be talking about work and how things are going. This is work because you need to know what's going on, and need to care about how your team is feeling and doing.

Cofounder management

It turns out that before founders ever have to manage employees, they have to manage one another. In fact, at the earliest stages of a company, when it's just two or three founders, bad management generally leads to the death of the startup. This kind of management also happens to be really hard.

Despite how important and difficult managing your cofounders is, most of the management advice I've seen is about managing employees or superiors. That's useful, but not at first. What I have read about this stage of management is structured as "treat this as a relationship." That's true, but also general enough as to be only somewhat helpful.

Managing cofounders is hard for a number of reasons:

  1. Initially, there are no clear lines of direct reporting amongst cofounding teams. One cofounder may be the CEO, but that doesn't mean that that person is the best manager or the best suited to lead engineering, sales, or product.
  2. Cofounders rarely have experience managing anyone or anything. They're learning at the same time as running a startup. There are a lot of good ways to manage, but they take time to learn and practice.
  3. Cofounders often think they don't need management because they're all on the same team, working towards the same goals.
  4. Startups are high pressure, and pressure makes people make bad decisions and lose their tempers. Small mistakes get magnified 
  5. Communication is much harder than you'd expect, even when there's just two people.
  6. Deciding to start a company from scratch with the goal of building a billion dollar business takes ego. As a result, founders often have large egos. With every success or piece of publicity, egos get inflated. Failures, public and private can deflate egos, and beat people up emotionally. That roller coaster creates tension, frayed nerves, and fighting.
  7. Divisions of responsibility are often unclear. That can result in turf wars, feelings of encroachment and micromanagement.
  8. Decision making in small teams of equals can be hard, especially when there are disagreements, which are often passionate.

There are many other factors that can introduce difficulties into cofounder management. Fortunately, the set of solutions is significantly smaller than the problem set. These problems stem from root causes which can be dealt with more simply than addressing the various expressions of those causes.

Open communication is the single most important factor in creating a good working atmosphere and provides the scaffold for everything else. It's significantly more important than cofounders liking one another. It is not enough for cofounders to agree to communicate, and generally inadequate for them to talk when the need arises. Cofounders need to establish regular check ins with one another to talk about issues at the company and with one another.

It is often helpful to have these types of conversations away from the office, especially once there are employees. Moving these conversations away from the office limits interruptions and also takes a lot of the psychological tension out of the conversations, especially the difficult ones. My cofounders and I were lucky here. We had a coffeeshop next door, a restaurant downstairs, and a bar across the street. Each came in handy, depending on the intensity of the conversation. We did, however, realize (a bit later than we should have) that disappearing from a tiny office too frequently during the day was a bad idea and hurt morale for the rest of the company. If you find that you can't have these types of honest talks with your cofounders, even out of the office, you're in trouble.

At the earliest stage of a startup, when an idea is morphing into a company, you'll probably have the first of your difficult conversations: you and your cofounders need to divide responsibilities and assign ownership of goals and the tasks that need to be accomplished to achieve those goals. Final decision authority has to be established for individual areas and for company wide decisions. Some of this will rest with the CEO, some of it with the head of product, engineering, or sales. Those roles might be filled by the the same person, but the responsibility flows through the role, not the person.

Remember that this conversation can get contentious if people feel they are being cut out of decisions they believe they should own. It is really hard to cede authority, but it has to happen in order to create a manageable strucutre.

Talk about these issues early, write down your decisions, and regularly review them.

Part of the reason you need to divide ownership of responsibilities early is to set expectations for yourself and your cofounders. Managing cofounders isn't easy, and you shouldn't expect it to be. When you're fighting over who has final say on product decisions, you'll discover just how hard it can be. One of my cofounders and I used to have shouting matches when I acted unilaterally on product without informing him. While he agreed that I had final say, I had failed to convincingly communicate my reasoning to him, and failed to set the right expectation of how product decisions would be made.

The truth is, finding out just how hard managing can be is one of the biggest shocks of working with other people. Having a reasonable set of expectations about it will help. Knowing, ahead of time, that you'll fight, get pissed off, and struggle, puts each of those events in a context that makes sense. Having strategies to work through each of those events means that those events (hopefully) won't destroy your company.

While you'll find some of your own strategies, it's a waste of time to come up with them completely on your own. Management is as much a repeatable and proven process as it is an expression of personal style. The process aspects can largely be adapted from general management literature, but I've always found it more useful to engage with a trusted mentor. In the best case scenario, this mentor is trusted by all the cofounders so that they're learning the same lessons and can use the mentor as an impartial arbiter when needed.

Finding a good mentor is tricky. Doing YC will give you access to some great ones, but certainly isn't the only way. Ideally, you want to find someone who has been through the situations that you're going to experience. You should also expect to need several mentors at different points in your career. The mentor advising you at the earliest stages might not be the person you want when each founder is responsible for dozens or hundreds of employees. The challenges you'll face will be different, as will the advice.

Even with this framework in place, managing your cofounders is rarely easy and invariably gets harder as the company grows and pressures rise. The pressures get amplified even more when things start going wrong, as they always do. You'll fight about important things, and you'll fight about seemingly inconsequential things. That's all ok and quite mundane. Just make sure you keep talking about it, adapting, and moving your company forward. Your goal shouldn't be to make your relationship with your cofounders easier. Your goal is to make the relationship manageable.

Someone else had your idea first

"I liked Jimi Hendrix's record of this and ever since he died I've being doing it that way..."

- Bob Dylan, on Jimi Hendrix's cover of All Along the Watchtower

Most people are very lazy. They don't want to take the time to think through new ideas or look at them in a new light. Once they've made up their minds about something, they don't change them. That's generally why most people don't come up with ideas for new or great things.

This is also true for many venture capitalists. In fact, it's at the root of a very common question that founders get asked: "Well, isn't so and so doing this?" To be fair, this question isn't necessarily sparked by laziness. It's also sparked by ego - the VC wants to show how familiar they are with the market. They say "Look! I know about things and there's someone else who had the same idea you had." The implicit criticism here is that, because someone else had the idea first, your idea is somehow worse.

I think part of the reason that people ask this question as a way of putting founders down is that they assume that startups are zero sum. That's an assumption born in certain models of markets, but it's completely wrong when looking at startups. Because startups create new value, the idea that someone else has done or is doing something similar to what you're doing often acts to broaden or prove the market you're attacking.

That's not to say that directly cloning another company is a great idea. If you have no differentiation and no original thinking on a problem, then you have to fall to one of two arguments: a) the market for a given idea is so large that there's room for multiple players executing well or b) the other company is so bad at executing that they'll self destruct. They're both potentially valid, but they're hard cases to make - especially at the early stages of a company.

Even though the question might seem dumb, it's one of my favorites. It's also a great question to get as a founder. I ask it of almost every founder I meet, because it's very rare to find a truly new idea. The answer I'm looking for is nearly always "of course someone else has tried this before." But that's not enough. The question begs for a deeper answer, one that talks about why, even though other people have tried the same idea, they're still leaving billions of dollars on the table. It's an opportunity to demonstrate depth of thought and originality. It's that framework of thinking and level of insight that makes greatness.

We're all communication hoarders

In April of 2004, Google announced that its Gmail product would give users 1 gigabyte of free storage. At the time, Hotmail offered users 2 megabytes and Yahoo offered 4 megabytes. I'm guessing I initially accessed my invite via PINE, and found the idea of using a full gig of storage for email to be crazy. Unsure what I'd ever do with all that space, I initially used it as a remote backup for my thesis.[1]

Ten years later, I have nearly 12 gigs of saved email - and I delete quite a lot. Like a family in a too large home, I hold on to messages I'll never need again for two reasons: 1) the cognitive energy to decide to destroy something forever is greater than the energy needed to put it out of sight for the time being and 2) Gmail's UX actively pushes me to archive rather than to delete. This principle extends across nearly all the communication mediums with which I interact. It is more difficult to delete pictures than to upgrade storage, more difficult to delete texts rather than keep them, and to accept social connections than deny them. In each case, my desire to save against the future wins out against the knowledge that, in all likelihood, the vast majority of what I save will never be useful to me.

This is a strange place in which to find myself. I don't like keeping extraneous items around. To be sure, part of that is a function of living in a NYC apartment with little room to spare. As opposed to my apartment, though, I can assume that my storage space is effectively infinite. Yahoo already offers infinite storage to its mail customers, and it's likely that the other players will follow suit over time. This makes sense when you consider two factors. First: how cheap storage space has actually become.[2]

Second: the data contained in my communication is more valuable to my email provider than what I'd pay for the space.[3] It is unsurprising that I'm given an ever larger shoebox to fill. With no obvious cost to keeping everything around, that's exactly what I start to do.

And that leads to a paradox. The more of my communication I keep, the less each piece means to me. It feels like I'm losing something as a result, even as I gain a trove with massive potential meaning. My wife's grandfather was in Paris during WWII with the US Army. In the two years he was away, his wife had their first child - an event he only discovered weeks later via mail. The letters they wrote one another are unbelievable historical artifacts that shape their and our understanding of them and the world.[4] Of all the things they could have saved throughout 73 years of marriage (and counting), they made the conscious decision to save these items. That decision is a key part of how we know their importance.

My kids and grandkids won't have the experience of reading letters that my wife and I have saved in the same way, because we save everything by default. It's entirely possible they'll have nothing since my email account will most likely be locked when I die. That doesn't mean that all this communication I generate has no value or meaning. It is hugely valuable, in aggregate, to Google and Apple and Facebook. They'll continue to have access to my information long after I die, and it will continue to feed their algorithms.

I don't properly know what I'm losing by gaining so many individual pieces of communication. I do know, however, that the pace at which we communicate continues to accelerate, and that the forms through which we communicate continue to evolve.[5] The ways that expanding body of communication gets mined for information are proliferating at the same pace, but so far, they're almost entirely geared towards the companies that make money off our data.

I think that leaves something on the table. There's a class of product yet to be successfully created that can sift through all of my communication, across all platforms, that finds what is actually meaningful. I don't just want the first message that said "I love you" to my wife, I want the letter or email that led to that conversation. I want to be able to find the text which, on the surface, was meaningless, but in another time I would have set aside as an important life marker. While I can manipulate my inbox search to find some of these things, I can't really find the meaningful things. Maybe Google already does this to serve me ads, but that doesn't really help me.

Then again, maybe I'm thinking about communication all wrong. Maybe it should only have meaning in the instant it is made because that's a better fit for our brains. Or maybe we haven't figured it out yet. That feels more accurate to me, and I'm looking forward to seeing what comes next.


[1] Pretty sure I wasn't thinking of "cloud storage" at the time.

[2] Chart courtesy of ZDNent. "Thailand Hard Drive Crisis" is my new favorite chart annotation.

[3] I've talked about this cost/persona data trade off before:

[4] We're lucky enough to still have Pops and Grandma Lil telling us stories. Grandma Lil also still has some of the perfume Pops bought her with bartered champagne and cigarettes.

[5] I recall being in Scotland in 2005 and being confused at the popularity of texting. It seemed so strange and foreign at the time. Considering the average American 18-24 was sending 3200 texts a month in 2011, I think I got that one really wrong.