What I learned doing a startup: part 2

This post comes from a series; read the series overview here.

As already mentioned Ground Control didn’t start as a SAAS business; it began out of my desire to have something resembling an agency. My attempts at actually building an agency had been somewhat faltering due to a lack of insight about how one actually gains business leads in that space. As such the promise of recurring revenue from a more modern style of software business seemed appealing. However a key thing to remember when starting a business is that if something seems too easy, it probably is.

The customer isn’t always king

One of my earlier customers had a great set of goals for their business. As such they seemed a perfect fit for the software platform I was busy creating and trying to sell around the country. At that stage I still hadn’t settled on a full SAAS model of actually building the software - each customer got their own deployment, their own database. This made the process of rolling out updates relatively tedious as each customer needed a new deploy running; I may address some of the more specific engineering challenges separately but for now I’ll stick to the business side of things.

What the deployment process meant on the business side of things was that it was relatively easy to promise different customers different features. Because deployment was already manual and slow, the extra requirement to load in and out different blocks of code for some key components seemed like a reasonable thing to do - it allowed me, working on my own remember, to be able to rapidly meet customer requirements without having to consider too much where the system was heading long term. Or at least I thought.

This likely contributed to the high engagement of some early customers - they were effectively getting the equivalent of an in-house developer at a fraction of the cost. However it had an adverse effect once the requirements of different customers not only pulled in different directions, but attained different financial incentives. As customers realised that they could effectively request bespoke work by offering ad-hoc extra payments above their subscription, the race was on for what time I had to work on the system.

What happened?

By the time I began to step back from full time work on the business I estimated I was spending up to 80% of work time on a few customers - customers who did pay more as one-off fees, but who in total still contributed less than half my revenue. This slide had gone on for so long that looking back I can see how clearly it damaged my ability to grow the business, rather than just keep it on a plateau. I was making, at times, just enough money with how things were going, but never enough to spend more on marketing or outreach to customers, or to hire other developers who could have tackled some of the increasingly complex technical knots I was tying myself in.

I think there’s a few important lessons to draw from this experience:

  • Firstly don’t take your eyes off the goal (and maybe point 0 is to have a goal in the first place?) It’s easy to let short term cash requirements, or an angry email from a customer, or criticism of what you’re building within your market, to distract you from the work and decisions that help a business to grow.
  • Secondly make sure you get a process in place early to establish how feedback from customers is handled. If you need to make promises when onboarding early customers then make them proportional to the long term value of the customer - doing 2 months full time work to get them to sign up when a monthly subscription to your product is £10 would clearly make no sense. Yes you need to care for customers, but you do that for all your customers by ensuring your business doesn’t falter because you divided your time unwisely.
  • Finally be prepared to have people be unhappy with you, complain at minor bugs that they thing are major, and even leave for competitors. The emotional attachment to early customers, to the revenue they provide and the praise they can give you can cloud your ability to make decisions about their true value to your business. Of course be reasonable, helpful, try and make it work with them - but don’t do that at the cost of the rest of your business.