Wednesday, 22 February 2012

Software as a Service

There are two ways to sell box office ticketing software: you charge a license fee, usually based on the number of users or something similar, and then charge an annual support fee on top of that, or you sell it as a service.

I'm selling mine as a service, because I've seen what happens over the lifecycle of a business that sells this sort of enterprise software based on license fees, and I don't believe it's in anyone's interests.

This is largely based on my experience at Artifax Software, but I experienced the tail-end of the same process at Galathea.

You start off getting your first few customers, selling your first few licenses. You get a nice pile of money in the bank, and you start to think, "Hey, this is going to work!". Once your sales pipeline gets moving, you feel like you've got enough of an income stream to start hiring staff.

So you've got a couple of programmers and a few support staff, and you're selling enough licenses to pay the wage bill at the end of the month. The trouble is, the more licenses you sell, the more support staff you're going to need to cover that number of customers. Now, your 15% annual support fee should in theory cover the increase in support staff you need, and the development costs associated with making the next version, but one day you blink, and before you know it you're using your license revenue to keep on paying your support and development staff.

At this point, it's clear that you've got to keep on making more sales in order to support the staffing levels you need to provide your customers with the service they expect. And the more sales you make, the more staff you need. And the more staff you have, the more sales you need.

It gets worse, because the accelerating need for new license sales means you have to start branching out into new markets. Which means more functionality: more developers, more complexity in the code, more bugs, more support staff. Now, every new sale means adding features to the product, not just using the intellectual property you have developed. But you make the sale on a promise of functionality because you need the money, and then you hope to catch up with the development before the next new customer comes along... adding more developers as needed.

What you get at the end of cycle is software that has been stretched to work in too many different markets. Revenue is drying up because you are literally running out of new customers to sell new licenses to. Your reputation for quality is going down because you have half finished functionality left right and centre, and your reputation for support is going down because the software is now hugely complex and you can't pay the support staff you need on the support revenues you're getting.

When I was at Galathea, my 5 biggest problems were a racecourse, a football club, a stadium, a theatre group, and an cultural institution. They all sell tickets, but they all care about totally different things, and none of those things worked properly because they'd all been done in a rush. The racecourse cared about fraud, touts, and access control. The football club cared about season ticket renewals. The theatre group cared about selling huge volumes of seated tickets at a steady rate. The stadium cared about selling huge volumes of general admission tickets in 30 minutes flat. And the cultural institution cared about memberships and the way that membership perks interacted with every other aspect of the system.

Software as a Service may look more expensive in the long run: if you add up all those per ticket fees over 5 years, it might be less that some up-front-licenses, support fees, and a couple of servers. But the SaaS business model means that your supplier isn't going to be panicked into making unwise sales  and rushed development, and it means they're going to care as much about keeping you on board in 5 years time as they are the day you sign on the bottom line. Which is what everyone wants.

1 comment: