Around 1990, a weird thing happened: IT departments started to experience the momentous soggy “thud” of Enterprise Licenses for Lotus Notes dropping on them from a great height; often procured in a corner office with no involvement by mere technologists. And another weird thing; IT management discovered that signficant jobs were getting done (and well, and quickly) using this weird language called “Perl” that wasn’t taught in any University courses and certainly wasn’t on Gartner’s radar. These can serve as canonical examples of the front-door (with management blessing) and back-door (with management oblivious) approaches to introducing technology. Let’s use our dozen winner and loser technologies to see if the level of management approval would have been much use in predicting their success or failure.
A Note on Methodology · The idea is, we take each of our winner and loser technologies and give it a score of zero through ten estimating how strongly some predictor (in this case “Management Approval”) would have applied to this technology. So if all the winners scored 10 and all the losers scored zero, we’d have a perfect positive predictor.
I cooked up these numbers entirely subjectively based on my own experience, and some may violently disagree. I’ll present them in a table, and follow up with a few paragraphs of justification. But hey, this is the Web: if someone disagrees, and does so in public, and is convincing about it, I can go back and change the numbers, and I will. On with the show.
The Table · For the “Management Approval” predictor, a score of ten would apply to technology that was featured in Forbes more than once; zero would be the stuff that staff got fired for using.
| Winners | score | Losers | score |
|---|---|---|---|
| SQL/RDBMS | 10 | OODBMS | 3 |
| Unix/C | 0 | 4GL | 10 |
| Open Source | 0 | AI | 5 |
| PC Client | 3 | VRML | 0 |
| WWW | 7 | iTV | 2 |
| Java | 3 | Ada | 8 |
| XML | 8 | SGML | 1 |
Discussion · The top score for SQL/RDBMS may be a little high, but I honestly don’t recall any pushback; maybe a bit of residual conservatism by the people who were (and in some cases remain) heavily invested in the old hierarchical databases. Weirdly, it seemed like all of a sudden there was a massive consensus that this was the way to go.
On the other hand, I’m pretty confident that the top score for 4GL is fair; I can recall the massive corner-office enthusiasm to this day.
And I’ll stand by the zero for Unix/C too; that was a technology that was totally brought in the back door and basically invisible to anyone who wasn’t a hard-core tech, until sometime around 1990 everyone noticed that Sun was making tons of money selling the boxes. That’s not just Unix, it’s C too; over the course of my career my bosses told me to use Pascal, then PL/I, then Ada, and at the same time more and more of the real work was being done in C. In a similar spirit, zero for Open-Source seems fair.
Both XML and Ada were welcomed at the top; but not as much as some of the favorites.
Almost all IT management and a lot of their bosses took a very dim view of the PC Client; but you can’t award a zero because line-level bosses all over the world who just wanted to print memos and do budgets faster were buying them like hotcakes in the face of this disapproval.
The WWW was pretty popular on Day One just because its interface was so intuitive; but Java, while it was still being presented as a Web page animation tool, drew mostly raised eyebrows.
Conclusion · It’s all over the map. For this sample of technologies, neither management approval, nor management disapproval, would have been useful as either a positive or negative predictor of success.
