Writing Software in a Feedlot

A lit­tle story: I’m from West Texas where cat­tle ranch­ing is big. In com­mu­ni­ties around the Pan­han­dle, there are feed­lots where cat­tle go to spend the last 3–6 months of their lives before becom­ing steak. Feed­lots can hold a lot of cat­tle. As you can imag­ine, the air qual­ity around feed­lots is pun­gent, to put it mildly. When the wind is blow­ing the right way, you can smell Here­ford in Amar­illo, 60 miles away. All that methane is a byprod­uct of the feed­lot, one that the peo­ple liv­ing near the feed­lot would gladly give up espe­cially if they could still get steak.

Funny thing is, peo­ple who work on a feed­lot don’t notice the smell, at least after they’ve been liv­ing there a lit­tle while. Like any con­stant stim­u­lus, our body even­tu­ally can tune it out.

What does this have to do with writ­ing soft­ware? Like the methane from a feed­lot, lots of com­pa­nies seem to view the soft­ware they pro­duce as a byprod­uct. Typ­i­cally the lead­ers of these com­pa­nies, like the peo­ple around a feed­lot, would love to get rid of soft­ware devel­op­ment if it meant they could still make money. These peo­ple tend to want to buy as much soft­ware off the shelf as pos­si­ble, even if it’s not a very good fit. They want busi­ness ana­lysts to be able to set up com­pli­cated cus­tom solu­tions using some sort of drag and drop IDE that never seems to quite live up to its promise. In the end, they wish the com­pany didn’t have to pro­duce soft­ware to make money.

When soft­ware com­pa­nies are run like this, you get an IT depart­ment that is mediocre at best with low morale and zero ini­tia­tive to improve the process of writ­ing soft­ware. Tal­ented soft­ware devel­op­ers go some­where they are appre­ci­ated, where their tal­ents can be used to solve actual prob­lems. When soft­ware is seen as a byprod­uct as opposed to an actual result, there’s no real dif­fer­ence between devel­op­ers since what­ever comes out will be the same. Everyone’s shit stinks, as it were.

A com­pany that treats soft­ware devel­op­ers like cows in a feed­lot will always func­tion at the level of the low­est com­mon denom­i­na­tor. The com­pany may not fail, but it will never really suc­ceed either. I recently left a com­pany like that. I don’t think it will fail. I just don’t think they’ll ever succeed.

3 Comments

  • and this is why it’s so much bet­ter to buy free-range, nat­ural, organic meat. ;-)

  • cleverpiggy wrote:

    ah, brett, that’s just the smell of money…or so all those amar­il­loans used to tell me…as for the soft­ware companies…my hus­band (a soft­ware engi­neer also) would most likely agree.

  • Scotch Drinker wrote:

    It was cer­tainly the smell of money, I just like my money to smell more like perfume.

Leave a Reply

Your email is never shared.Required fields are marked *