How Not To Build A Product (Startup)

I'm on my second serious go at bootstrapping a start-up.

The first time I (we) tried to build a company, we failed. Badly. And I learned so much from the experience about building a product; painful lessons that I want to share with you today.

You can interchange "startup" with "product" if you're working for a BigCo, and the lessons will be the same, really. After all, how different are these journeys? Besides the latter having capital, resources, and a vast network of aligned colleagues.

Anyways, what are these lessons?

Start-up Number #1, Ajtima

I was recruited as a technical co-founder by a friend to partner on building a marketplace that would change the world.

Or so we thought. I really believed, and continue to believe, that we could have made a significant difference by building a Fiverr clone, with the fundamental difference of focussing on specific, under-served regions of the world. One of my mentors advised and said,

You don't have to create a new product, you can execute on a different version of another product and provide value. Think BMW vs Merecedes vs Audi, and so on.

Not exactly sexy, but, wise-words!

And thus, Ajtima was born. I hired another trusted developer, and we spent 6 months building an infinitely scalable, best-practice, polished, Fiverr clone for the Middle East.

Amazing right? No. Who gives a shit.


No Seth, it's not cool.

We built a product in complete isolation with no UX testing & no market validation.

The marketplace was supposed to serve freelancers and customers in the Middle East, and we didn't stop to ask a single person from the Middle East what they thought about that.

But wait, there's more! Instead of building the project lean, and pushing it out to the early to test our untested hypothesis, we spent 6 long months "perfecting" it before we thought about how to get people to use it.

I'm talking automated tests and a robust CI/CD pipeline, re-usable internal libraries, real-time native feel in a browser via Subscriptions and Redux pattern, SSO, the works!

This was the outcome:


The world's saddest play place

So what did we learn? Two sides to it.

From a business perspective, ensure your product fails by building it:

without direct user feedback
without aggressive focus on product market fit
in a blackbox

From a technical perspective, what did we learn?

SO MUCH. But that's a different story, and a different blog.

Long story short, though Ajtima will haunt us onine forever, thanks to cheap PAYG software, the project is very much dead anad without momentum, and we've all but since moved on to different ventures.

Looking forward, how were these lessons applied?

Start up #2, Serv

The next time I was approached to be a technical co-founder, I knew a few things.

Don't build something people don't want
Don't build something people don't want

Can you guess what #3 is?

So I told my co-founders, that I wasn't going to write a single line of code until we had our first customer.

That's right.

But I made them a promise, that once they found a customer, a real customer, I would build them a PoC in 2 weeks.

That's right. I'm an idiot.

But my intentions were good!

Don't build something people don't want
Build the product lean to start, think MVP
Let your customers help you find product market fit

A few weeks after our initial meeting, I get a call from my co-founder, who says they had pitched the idea to a restaurant, and the manager was excited to try it out. So with that, the clock began ticking.

And thus, Serv was born!

Equipped with my experience on how to build scalable products thanks to Ajtima, I went to town. We had numerous meetings to reduce the scope of this incredible, grand, world-changing idea they into a minimum valuable (not viable!) product.

Once we had something simple that would still perform the basics of what Mr. Customer was looking for, I went to town, and 2 weeks later, we had a working prototype that we handed off.

And oh boy were there bugs! But did that stop us or Mr. Customer from using our MVP? nope.
And on top of that, we were lucky enough to have a very vocal first customer. He shared every gripe, concern, pain point, headache. We were very fortunate!

We used all that feedback to improve the product, from a bare-bones MVP, into a somewhat respectable product.

When we onboarded our second restaurant, we kept that process going. Now we were fortunate enough to have TWO vocal customers who fed back all their comments and suggestions.

Fast forward to today and we have a product that is:

iterated on quickly
evolving from MVP to product market fit alongside real users

Will it go the mile? Who knows, but add your email to my blog so you can stay up to date on the latest :)

Thanks for reading.