Recently between contracts I tasked myself with building a startup that would help digital agencies better manage their staging servers. Despite reading The Lean Start Up and the Four Steps to Epiphany I managed to witness myself, in real time, making every mistake in the book(s) as well as a few extra mistakes I’ll outline in this post.
Deciding on a name, registering a domain and designing a logo
In a rush to make the dream more real I went about deciding on a name (‘StagingBox’), registering 2 domains (stagingbox.io and staging.sh) and designing a logo (shown on the left). I even designed a neat favicon for the website. I approached starting a business like a web developer would. For example I planned out the architecture of the app and invented a glossary of terms for different models. A ‘Box’ was a cloud server instance, a ‘Client’ was one of the apps customers and a ‘Host’ was an apache vhost (or in the future a Docker instance, because Docker is so hot right now). So why is all of this bad? Well - because I hadn’t even worked out who my customer was, the minimum feature set they needed, whether they really cared about the problem and whether there was enough money in it to sustain a business. Creating the branding and architecture gave me a way to fantasise about how awesome my business was going to be without actually knowing what it was going to be - which is exactly how every single piece of Vaporware ever made probably started.
Not knowing my customer was actually my worst client
My product was aimed at digital agencies who I’ve worked for in the past 5 years who have had pretty dodgy server infrastructure. Traditional digital agencies tend to pitch hard to win new work and spend a lot of resources generating visuals and fast prototypes to win over contracts from their competitors. When it comes to infrastructure on the other hand, especially for staging servers - which is what my app was all about - digital agencies tend to spend as little as they possibly can. I forgot to ask myself “do I really want to work with digital agencies who don’t invest in their infrastructure enough to go with proper PAAS solutions?”. In fact, for the past few years I’ve been avoiding contracting for those agencies and now I was trying to get back into business with them!
Making a pretty prototype
I have to say, I made one sexy prototype. It has some really great performance, it’s scalable, unit tested, responsive, animated and yet it doesn’t really do anything. It does interact with the Digital Ocean API to create new servers but beyond that it’s pretty much useless. I was so caught up with making the fantasy real I created what basically amounts to a cardboard hollywood set. I’ve spoken to successful startup founders recently who said “Oh man you should have seen how crappy the first version of our product looked” and yet their MVP actually proved the problem, proved their solution and functioned as a way of getting further feedback from clients. As a developer I drew on my experience with corporate clients and created a perfect developer experience, great UX and well structured code. I should have closed all my blinds, stocked up on frozen pizzas, given myself 4 red bulls, no more than 24 hours and gone after a single killer feature.
Keeping it private
That tired line of “I can’t tell anyone my idea because I’m worried someone would steal it” is absolute bullshit - what it actually means is “I can’t tell anyone my idea because I’m worried they’ll criticize it and I’ve been dreaming about it for 6 months and it’d destroy me if somebody said it was bad”. I’m not saying go around telling everybody the intricate workings of your app but at least go around telling people what the problem you’re trying to solve is. If I went to agencies and said “hey, I’m working on an app that’ll solve your problem with managing access to staging servers and allow developers to easily create new ones without SSH access’ I might have got 1 of 2 responses. One response might be “That’s great, we really have an issue with that - we also have an issue with XYZ”. Another response might be “What are you talking about? That’s not a problem my agency has. Who are you? What are you doing in my garden?!”. Talk openly about the problem you’re trying to solve, or at least do a bunch of ‘problem interviews’ with potential clients.
Assuming that a lean, one man B2B company was a good idea
I built my company (or at least the dream of it) on the basis that if I kept the product very lean I could run the whole thing myself, take all the profits and work from anywhere, anytime. Great in theory, but even if the product managed to sell itself without salespeople what kind of digital agency would trust a one man show? What if I got run over by a car the next day, would thousands of agencies staging servers be floating in the abyss as the company dissolved? No right minded business is going to trust a one man show to manage all their staging server infrastructure. I was going to have to hire people, and I just didn’t care enough about the business to do that. Consumers on the other hand will make one off purchases on non-essential and almost disposable goods on a regular basis, and the same applies to digital purchases - so the one man operation works much better for B2C than B2B.
Anybody want to buy my domain names and steal my idea? I'll throw in the logo for free!