Tuesday, November 6, 2007

In defense of developing enterprise software

Developing enterprise software is not what many claim - easy. The products created at Workbrain, which has been acquired recently by Infor ( the largest private software company you've never heard of ), are used by some of the largest companies with employee bases of well over 100000( airlines, retailers, governments, manufacturers ). Performance under simultaneous load from reporting, scheduled tasks, and UI access points is paramount and heavily tested to maintain best usage.

I can't completely argue against Khoi's post, but yes, design and usability are continuously improved. When starting a company or product, the first few months will typically see explosive growth, followed by well formed growth, then normal progress. As development matures, more care is needed to minimize the risk of regressions and satisfy the massive population of current users. Separating the core product from innovative new product ideas is a great strategy to keep enterprise software companies feeling new, nimble and competitive - without that, they typically end up giving their products and services away and billing for consulting or documentation( aka IBM ).

Jason nailed the difference in how products are managed at enterprise companies - the users of the system are not those that the software is marketed to and many times not who makes the buying decisions. I wish it wasn
't the case but how can you sell enterprise software to individuals ? The good news is that we cater to the users through incremental revisions, feature additions and formed a working group where customers can basically vote to influence a percentage of our direction.

Each time I re-read Paul Graham's essay on "why to not start a startup", I laugh hard enjoying the elitism, but also get pissed because my technical reputation is being eroded. Perhaps many, if not most, business problems in the enterprise arena are easily tackled relative to recent startups, but that is not convincing. Many problems are far too complicated to begin without significant research, as opposed to the ability to craft a website that ranks recently viewed materials or allows you to enter your task list. I
'm not saying that developing for the enterprise is dreamy - but don't compare it to old school crap.