But, but, but

After I wrote "The Coming Storm" Sandhill placed an even more provocative article, "Is Enterprise Software Doomed?" . This stirred up such a frenzy of objections that Sandhill felt compelled to place a collective rebuttal "Software's Sky is Not Falling" . In reading this I felt that there were some important points missed. So I left them as comments on a blog somewhere out there. Then I thought I should write them down here, as I think that they are points I need to take on board in my journey through the business of software.

Where are the developers in this debate?

Why? Well, enterprise applications are complex. The source code is generally poorly documented. The body of knowledge really exists in the developers heads. Even with good guides and well documented code it takes a great deal of time for new developers to get up to speed. To understand the applications limitations and abilities. And to understand the clients needs.

This is why in the open source world there are very few examples of one product being successfully taken over by another company. And when it does happen it is generally lead by a developer revolt.

This is why the outsourcing of existing enterprise applications is a rather dodgy idea. It is also why cutting costs by replacing those expensive Java developers with cheaper Oracle Forms developers is a bad idea. Don't laugh: this actually happened at a company I worked for. The product is now dead :-(

This is why when buying or investing in a software company I would leave the balance sheet till last: I would look first at the history of the developers and their relationships to the products they work on. As pointed out, "Open Source" is simply a label that covers many business models. But a model in which it doesn't matter if you download and use the code free of charge. A model in which it doesn't matter if competitors have access to the source code*. And why? because these companies have their developer well and truly captured...

*If this isn't a scary thought for incumbents in the enterprise application space, I don't want to know what keeps them up at night!

But the customer already has the software they need...

I worked on a legal practice management system for quite while. And whilst I was there sales slowed. Towards the end of the last century they eventually almost stopped. Why? Because we ran out of people to sell to. All the law firms now had the software that they needed.

The market had changed beneath our feet. The only new customers were ones that became disaffected with their existing vendors. Our sales and marketing team became a dead weight. One that drained funding from the products development. But it was a dead weight that conventional wisdom dictated we need.

Our sales team couldn't be there when someone else's customer slammed down the phone with a "Damn that vendor! Time to find another..." But Google can. One short search, one quick download or trial run, and the customer is gone. And similarly for new firms: our sales team couldn't find all the new lawyers hanging out their shingle for the first time.

But Google can. One short search, one quick download or trial run, and the customer is signed up. So really, the enterprise market has become one in which the only new customers left are ones that are new to the market (small) or disaffected. This point is discussed much better than I can here...

The existing vendors are captive to their customers

Anyone interested in the future of software must read the "Innovator's Dilemma". The most important lesson in the book, to me, was that incumbent companies receive most of their value from their existing customers.

They are thus locked into their customers requirements and models. It sounds obvious, but this also seems to be so often overlooked. This is why the existing vendors, who should have all the running, will struggle in the "coming storm". They will find it very hard to make the required changes.

But:

The most interesting point is that "we are today seeing the emergence of sustainable large-scale production projects that don't rely on either the pricing system or management structure". This is from Nicholas Carr's blog at http://www.roughtype.com/archives/2006/08/open_source_as.php

How are enterprise software companies going to cope when one of these production systems get going in their space? It is only a matter of time.

9:13:18 PM