Saturday, February 22, 2014

The Mid-sized Quandary

Being a medium sized business is a little like being an adolescent: You're old enough not to be a kid any more but you're still not old enough to drive. And driving really is the measure of adulthood isn't it (at least in the car culture we have here in the US)? For a business this size really presents a problem. You're big enough to have some enterprise needs, small enough to still be agile, but you're not big enough to bend vendors to your will.

I was reading an interesting article shared by +Bob Martens about Kiwibank choosing to move their core banking to the cloud with the SAP behemoth and I felt like I was looking into a mirror. Between the debates of "to cloud or not to cloud" and how should we handle our core systems and even more scary, how do we handle legacy systems that we rely on for day-to-day operations? The burden of legacy core systems is even heavier now with "The Cloud". Not too long ago the legacy system was tolerated long beyond its useful lifespan because of the expense to create a new system to replace it was just too much to stomach. Now there's vendors falling all over themselves to offer small to medium sized businesses a cloud solution to replace these legacy systems.

The siren song is all too familiar: "Tired of all the cost?", "Want to get rid of those expensive employees who maintain your legacy?", "What happens if it goes down? Who's left to save you anymore?", and so on.

Worse yet are the vendors who put lipstick on the pig and retrofitted their legacy systems into the cloud and slapped a web interface on it and offer it as a side step into cost savings and reliability without the need to ditch your current legacy. There you are, still stuck with the same limitations that the old RPG system running on an ancient release version of OS/400 because that's what it still is behind the scenes.

The problem is that cloud all to often delivers the one-size-fits-all design. It's the new mass production. Start small, grow huge, sell off to another dreadnought-class conglomerate, lather, rinse, repeat. Its very effective at raising money for investors but terrible for the business customers that rely on the cloud services. And that lack of flexibility to meet the individual needs of a particular business or even business sector doesn't help the organizations that need the flexibility most of all.

Mr. Wiggs is maybe overly optimistic about the do-it-yourself approach. If you asked a local team of developers to create a new bespoke core system for a bank you will almost certainly experience failure in its most pure and epic form. Double that failure level if that team has no experience in banking or financial cores. Some specialty systems can be a trouble to develop no matter who does the work. The point I like most about his perspective is that you shouldn't rule out the locally sourced talent, do-it-yourself approach.

Not too long ago (late 90's early 2000's) small local dev teams made a huge splash in the world writing software for companies in need of solutions. Those were still the pioneering days of computer systems but despite the complexities of modern technology and regulation there's no reason we can't have more of that today. Perhaps the patent landscape is too filled with perils to allow that to happen today? I don't know but if we want to grow small and medium businesses, one-size-fits-all does not.