My Biggest Mistakes - Part 1 of ???

Right now, I find myself in a position at Extra where people are looking to me for help in making good decisions. I'm working with an amazing team, and I'm honored that they value my thought processes.

There's a quote from Love and Monsters, though, that rings very true: "Good instincts are learned by making mistakes." I've made more than my share of mistakes.

As painful as those were to live through, I often find myself having a relevant experience that guides what we're going through today. Maybe these will be useful stories for others--and who doesn't love a good train wreck?

I started my first company, Blue Sky Labs, with two cofounders right before I graduated USC. Our idea was to bring Voice-over-IP (VoIP) calling to college students, long before anyone had ever heard of Skype or Vonage.

At the time, college students were paying $0.10 per minute for long distance calls, while our dorms were wired with 100mbps internet (for comparison, my current cable connection is till only 50mbps). No one really had cell phones or laptops, though, so the idea of making a call over your computer meant being stuck at your desk, speaking into a computer microphone and having the other person's voice blasted over your computer speakers. (Bluetooth wasn't even a thing yet, so if you wanted to move to headphone, you had to climb under your desk and plug them into your sound card.)

We thought that if we could come up with a way to use any phone to seamlessly make both normal and VoIP calls, college students would adopt it like crazy. We did the initial surveys, and the feedback was universally positive--especially among those who had long distance boyfriends or girlfriends.

Our idea was a box you'd plug into the USB port of your computer and the phone jack in your wall (no cell phones, only landlines at the time!). Then, you'd plug your phone into that box. You could receive calls from either the normal or VoIP line. By default, the phone would make normal calls--but if you dialed a prefix, the box would place the call over the VoIP line. We'd charge for the box and a monthly and/or usage fee for calls made to regular phone lines. Calls to other on-network VoIP lines would be unlimited and free.

My best friend and roommate had studied electrical engineering and designed the actual box. Our other cofounder, who had studied computer science at Stanford, led the team making the desktop software. I was the "business guy", which meant I was the CEO, as well as leading marketing, finance, fundraising, recruiting, and everything else.

We raised a seed round of $300k from friends and family (led by my parents) and got to work recruiting. I was 22, and the only real job I'd ever had was at Disneyland over a single summer. I'd studied biomedical engineering and taken a business minor, but I had no idea what I was doing. I also had no network to tap for useful advice (lots of unhelpful advice though!)

It was near the end of the summer--late August or early September, and we were about 6 months into our project. Though we had been incredibly scrappy, we were nearing running out of money.

Back then, there was no AWS or remote work. Step 1 to putting up a website was buying a bunch of servers and signing up with a colocation facility to host them in a rack. We had to have an office or there was nowhere to put our team.

We had also hired too fast. Though everyone worked cheap, we were building a hardware startup that required actual prototyping of printed circuit boards. That took time, and our marketing people on our payroll while just waiting for the product to be ready to sell (we'd had a few delays in product development,  so they were blocked).

Then we came to the fateful moment.

My cofounder came to me to report on the status of the latest prototype. It worked! But, there was a pretty nasty echo from time to time, and he can't compensate for it given the current design. If we wanted to eliminate that echo, he'd need to revise the printed circuit board and add another chip (a digital signal processor) that would be able to cancel out the echo.

What did I want to do?

I still remember the moment of the decision with a strange clarity. I stood there, staring at him in our office on the 6th floor of the dumpy old Wilshire Grand Hotel. The carpet was way too blue under our feet, and the eggshell white paint had somehow turned out light blue too.

I knew it was going to be make or break for us. If we shipped with the echo, the calls might be so bad that we'd die from negative customer reactions. If we tried to fix the echo, we ran the very real risk of running out of money before we shipped any product.

I figured that we only had one chance to make our customers happy, but we could use the working prototype to try to raise more money. Let's fix the echo.

You may have guessed how it turned out. We burned the rest of our capital iterating on the design, got to a working prototype, but were unable to raise the money we needed to launch.

Some of the outcome was undoubtedly due to the fact that we were trying to raise money in the summer of 2000, during the aftermath of the dot com bubble bursting. After all, the fundraising picture probably would have looked a whole lot different the yeah before. I had no ability to put the events of the time in context, and I couldn't imagine the convulsions that were spasming through the venture capital community at the time.

By the time of my decision, there was likely no way forward for us. Self-funding a hardware based startup was completely beyond the abilities of 22-year-old-me.

The lesson didn't arrive right away (it actually took me many repetitions before I got it), but eventually, it sank in. Despite the way that one moment was seared in to my memory, that wasn't the mistake.

The mistake had happened long before, and had kept happening until the eventual shuttering of the startup. My memory is just of the most indelible experience of the mistake--the moment when the weight of the mistake hit me. We were trying to build something "perfect" in a vacuum.

If I had gone the other way and used the rest of our remaining capital to ship the imperfect product, there were three potential outcomes:

1. Customers bought it and loved it, despite its inevitable flaws
2. Customers bought it and hated it, because it didn't work as advertised
3. No one bought it in the first place

Whatever may have happened, we probably would have still run out of money given what was going on in the world around us. The real problem is I will never know if I was right about the market opportunity.

Yet, in the beginning, that's the only job of a startup: to validate the hypothesis about the market opportunity. All of that effort was, in essence, wasted. We could have just spent that time surfing and ended up in the same spot (from an entrepreneurial perspective).

Now, when I'm asked, this is what I tell aspiring entrepreneurs:

Until you achieve product market fit, your job is not to gain customers or earn revenue. It's not even to build a product.

At the start, you have only one job: validate your business hypothesis. You do this by making a list of all the hypotheses as to what could kill your business hypothesis and invalidating those as quickly and cheaply as possible.

As you're doing this, you'll at some point you have de-risked your situation enough to move into the next stage. But until then, don't lose sight of this. In the beginning, progress isn't measured by the size fo the business. It's not in any way about "growth" or "traction". It's all about how much you've reduced the uncertainty around your business.

For us, back then, the hypotheses that could kill us were (notice how they're all phrased as bad outcomes--we're seeking to invalidate them):

1. College students wouldn't spend $100+ on a box to let them make VoIP calls (as opposed to using other services from their desktops)
2. We couldn't charge enough to pay for the costs of connecting VoIP calls to regular phone lines
3. The call quality of VoIP calls was too poor to retain customers
4. Colleges wouldn't find a way to block our calls to support their phone monopoly
5. We couldn't prevent better funded competitors from taking our niche

We spent $300k, several months, and an immeasurable amount of emotion, but in the end, we invalidated exactly zero of those hypotheses. In startup terms, we got exactly nowhere.

We could have invalidated hypothesis #1 without raising any money at all--just presell the boxes! #2 could have also been partially invalidated during a presale process. Having 100 or 1,000 or 10,000 orders (that we absolutely couldn't fulfill) would have been more meaningful progress than any working (but unshipped) product. Having those orders would have made it far easier to raise money too.

Now, there are several books intended to teach founders to think this way. Two of my favorites are The Lean Startup and The Startup J Curve. But back then, all the books were about how the internet had changed the world and how dot com companies weren't bound by the same business gravity as other companies. I was 22 and flying totally blind (worse than blind really). It's no wonder that I didn't figure it out.

Yet, over the next 20 years, I've found that the lesson hasn't really taken hold. Lots of founders fall in love with their product or don't want to ship something unless it's perfect.

I get it. It's scary to ship a product--especially an imperfect one--and get the brutal feedback that comes with direct contact with the world. It's hard to dream up a sparkling new future and then systematically seek to crush it. It's only through the pain of repeated wasted efforts that I'm able to stick to the better process.