John Sumser, my favorite luminary for the past 10 years (hard to believe that John and I have known each other that long), posted a comment to my post titled “Good Training, Bad Software.” John is justifiably famous for making statements which cause spontaneous reflection, even while the statement itself seems crazy at first blush. True to form his comment prompted a strong initial reaction: “This is crazy even for you John!”
My exceptions to John’s comments are several-fold. Saying that ERP software isn’t bad is erroneous by most measures: information architecture, user interface, cost of implementation, cost of maintenance, interruption of business, failure to be able to accurately calculate ROI: you take your pick, ERP software is bad software.
And software is clearly not “dead.” John seems to be taking the Nicholas Carr “IT Doesn’t Matter” point of view. That debate continues to rage, and not many minds have been changed. But most people agree that there is a lot of life left in software. John’s assertion that “the ERP’s have discovered the applicability of the limits of automation” is inaccurate. There is still plenty of work to do. Even more, there is plenty of work to do in inputs (including speech and handwriting recognition), outputs (including real-time, real-life voice response) and search. Software is not only alive and kicking, it is still in it infancy.
But as with most John Sumser wisdom, his comment got me to thinking. Assuming that I understand what he is saying, I have to agree with the general thrust of his argument (though not his particulars). The present software model is indeed terminally ill. It may take 30 years to kick the bucket but it is without question that the days of installing large software applications behind the firewall, developing large staff functions to maintain them and running the business by the fiat of whatever package the IT department agrees to fund will be a decreasingly viable model. The Salesforce.com model (which I will be writing about in a future post) is the direction that the software business is headed. The "software as service" model is very dependent on a strong services lead to the services / software mix, with the software forming the foundation of the knowledge capture and development offered by the services.
This doesn’t mean that there isn’t anything left to automate. Far from it: there will be more automated in the future than has been automated in the past. And it certainly doesn’t signify that the concept of software is dead: there will be more software innovation in our future than in our past, and we will be more dependent on software in the future than at anytime in our history. But it is an acknowledgement that the way that business people think about software and its value is changing rapidly. So kudos to John for possibly arriving in the right place for the wrong reasons. Possibly.
As always, comments and questions are welcome.