Monday, February 14, 2011

How long can your software remain broken?

I was very pleased this morning to see Seesmic roll out a bug fix for its Desktop 2 app (Silverlight).

I'm an ex-HootSuite user—one that was quite satisfied with that platform's offering until their commercial release completely missed the price point. I first became a Seesmic user by way of their mobile app for Android. It's a genuine pleasure to use, particularly if you're managing multiple accounts (and I manage over half a dozen). I'm not completely sold on Seesmic Web as a HootSuite web replacement, but on my Windows PC there's no question: I love Seesmic Desktop 2. I tested a slew of alternatives including the almighty Tweetdeck before I found myself hooked on this slick, functional, beautiful product.

Naturally, it was fairly irritating when I updated the SD2 software on January 18 and found that my ability to share URLs had been broken. Apparently, URLs shared by Seesmic's automatic shortening were unaffected, but for people like me who use their own preferred shortener and paste in the link, these were stripped by the app in the process of posting them to Twitter. Maddening. Embarrassing, too.

I can't fault their customer support, @AskSeesmic, for their willingness to help. At least three personnel managing that account contacted me about my problem, even when I didn't go out of my way to direct a complaint to Seesmic. Clearly they're scanning constantly for mention of their brand. But I'm very disappointed with the length of time it took to offer a fix or a workaround.

After remaining broken for two weeks, I honestly wondered who was more foolish: Seesmic for allowing the problem to persist that long, or me for holding on foolishly with my own self-discovered workaround—double-posting all my URLs to ensure that one would make it through.

So what went wrong? I don't know, but here are some of the options that I would have found acceptable:
  • A temporary re-allocation of programming resources to fix the problem in a reasonable amount of time, say, one week.
  • An emergency update containing a warning about the buggy update, and an option to rollback to the previous, working version.
  • Better communication. If the software developers knew that the URL-dropping problem could be avoided by using SD2's internal shortening feature, they could have mentioned this to @AskSeesmic, who might have tweeted me. They could also have emailed me directly, seeing as I filed a bug report and exchanged several emails with their patient and helpful rep Christine Wong.
Instead, none of those things happened. Inexplicably, I hung on to my broken software  until I received an update today: four weeks after my software was bugged.  That's just too long, particularly for a company in the process of investigating the business value of their product. Last week, Seesmic sent me a survey that seemed to be reaching for just that. I was initially pleased to see that they'd written, because I hoped it was a helpful response in line with one of my wishlist items above, but... no.  And instead, it looks like they may be following in HootSuite's footsteps.

Here's my take-away from this experience. Another list of things that shouldn't need to be said, but I'll say them anyway because nothing about this last month with SD2 has made sense to me:
  • Leaving free users with broken software reduces your total install base. I'm certain many fled to Tweetdeck in the last month, because there are few reasons not to like it.
  • Leaving one piece of software broken for too long a time undermines user trust. If you'll do that with your Product 'X', what's the chances you won't do that with Product 'Y' or Product 'Z'?
  • Leaving any piece of software broken for a significant length of time undermines your firm's credibility and appearance of competence. Who's running the show there? How many projects are being juggled concurrently? There's either poor management, or too many offerings.
  • It's irrational to expect that free users who have observed this management or endured these problems will convert to paying customers.
I'm sticking with Seesmic, but they have some trust to earn back and some more thinking to do before they venture into business.