Don't Outsource Your MVP

If you're starting a tech startup, please don't outsource the development of your MVP to a third party. Find a technical cofounder.

I know I’m not telling you anything you haven’t heard before. But please listen. I've heard too many heartbreaking stories from founders that tried to do it the hard way.

The founders that funneled tens or hundreds of thousands of their own money to a dev shop, without a product that works. Bled dry, a few thousand at a time, hoping that this time, they'll fix all the bugs and ship critical features.

The founders that gave up almost half the equity in their company to launch an MVP. Only leads aren't converting, and maybe they'll find traction if they add another feature or two. But the project's "done", so founders can't get any more time from the dev shop. And new investors won't fund the company because the cap table's all screwed up.

A Confession

Many years ago, I started a custom dev shop. Pretty quickly, my entire client list consisted of nontechnical founders who were using me to develop their MVP.

I loved startups, so was eager to work on these types of projects. And I was pretty picky about which ones I took on, so only worked on products I believed in. I hit almost all my deadlines, and charged way less than most of the competition.

But a couple years in, I shut down the dev shop to work on Formstack full-time. I transitioned all my clients to other firms or freelancers, and applauded them from afar.

My first startup client brought all development in house shortly after they launched. They raised a few $M from angels and VCs and later exited to a Fortune 100 company.

My other clients met less desirable fates. None of them got much traction with their MVP. I don't remember any raising that much money from outside investors, if at all. Some had enough capital to fund a few more iterations of development to give the company a shot. All those businesses are now long dead.

I've felt guilty about this. I put a lot of heart and sweat into the work I did for them. We built a good relationship while working together, and I wanted to see them succeed. How was I able to make Formstack successful, and not do the same for them? How much of the blame for the failure was mine?

But as I reflected and listened to other founder stories, I realized launching with outsourced development is a near-insurmountable uphill battle. There wasn't anything I could've done differently for my clients. The odds were stacked against us.

You'll Move Too Slowly

It's critical, as an early stage startup, for you to be able to iterate quickly. I don't care how perfect your idea is — you'll need to make frequent changes in response to your product touching real-world customers. You'll more likely be struck by lightning than not have to make changes to your MVP to find traction.

The dev shop's going to be working on many projects in parallel. So your 40-odd hours of resources will be spread across several weeks, not one. Or, they're going to assign resources serially — once your project is "done", they're on to the next one. And good luck getting resources until that one's complete.

Nothing beats working with a cofounder or team member side-by-side, in real time. Critical bug that's preventing new users from signing up? Fix it now. Metrics looking bad and you need to overhaul the onboarding process? Drop everything else and crank away to ship by the end of the week.

Some dev shops will be able to respond quickly, but only for a while. Their business model is always working against you here. For every change, you'll need to go through the process of scoping out requirements, bidding, and scheduling out resources. And as a founder with time working against you, that'll bring lots of stress and frustration.

You'll Burn Too Quickly

If time isn't against you, cash is. Working with a dev shop will always be far more expensive than working with a technical cofounder or team member who's doing the work.

Again, there's no getting around it — there's overhead baked into the business model of the dev shop. They need to make money. For every $10k you pay them, I guarantee that you can find a great full-stack engineer who's eager to work on something with you for a fraction of that price.

And what most first-time non-technical cofounders don't realize is how much work comes after the launch of the MVP. The dev shop may even say that to you, but it's easy to let optimism cloud your judgement. You'll spend a tiny percentage of your total development costs on the MVP. Where it gets really expensive is afterward, when all those bug fixes and iterations start adding up.

You might get into this trap where you blow through your bankroll before you're able to demonstrate traction. Or even if you've already raised some investor money, you might funnel all that to the dev shop before hitting important milestones. In either case, it's practically impossible to get the next round of funding.

In the end, this ends up costing far more equity in the long run. It's cheaper to find a great technical cofounder and give them 20-50% of the company on a vesting schedule.

You Won't Build The Right Thing

The result of all this is that you likely won't get a chance to build exactly what you need to put your startup on the path to success.

It's not only the lack of speed and limits of cash that lead to this. It's the intangible of working with a cofounder that's bought into the startup as much as you are. Someone whose hopes and dreams hinge on making this work. Whose anxieties and fears revolve on making sure this thing doesn't fail. Someone that's thinking about your startup in the shower, on the drive home — just like you are.

You're not going to get this from a front line coder at a dev shop, no matter how skilled they are.

Who's going to put in the effort to find effective ways to cut development time? Who's going to search endlessly for elusive opportunities for 100x outcomes?

Some Exceptions

There's no one way to build a successful startup, and exceptions to every rule. I'll even bend my "have a technical cofounder" rule for a few exceptions.

Is your startup not really a tech startup? If custom software isn't really core to your business, then none of this applies. If you're creating a shoe company, work with people that can help you build the best shoe possible, whether or not they can code.

Do you already have a lot of your own money that you're willing to spend to get this startup off the ground? And you're unusually averse to the idea of hiring a half dozen engineers with that money instead? Then go forth and bless a local dev shop with your riches.

Do you have experience launching a successful product within another company? And done so as an engineer or product manager? And you're only planning to hire a contractor or two while you do a lot of the core work, not outsource completely? Then none of this applies either.

What Do You Do Now?

I know it's hard to find a technical cofounder.

Maybe you already hired a dev shop, and they're promising to deliver exactly what you want, on time, and within budget. What do you do?

Fire them.

You need to ignore sunk costs. Find a technical cofounder, or take what money's left and hire a "founding engineer" to work on it with you full time.

Would you start a restaurant without a chef? Then why would you start a tech company without someone who can build your product?

It's hard enough building a company. Give yourself the best chance possible for success.

Get notified when I publish something new: