[DELETEAgile Software Development] Rapid Software Releases: A Critique
April 4, 2007
It’s all the rage – build stuff quickly and get it out there quickly, see what plays, what doesn’t, and change in the next release cycle. Yes, I get that things need to move more quickly than the old 6-8 month design and development process (I’m talking Web sites).
But as a somewhat early-adopter user, it’s starting to get to me. When Tabblo released a new photo-book creator in “beta”, I had to first email them to ask what was beta. Because if I was going to spend hours uploading and arranging the photos in their neat little application, it had better work. (In this case, it did brilliantly, which is why it’s so tempting to try new things every now and then.)
When I tried to use Upcoming.org‘s private event functionality, because I’m too cool for Evite, the experience was abysmal. Suffice to say there are strangers who think they were invited to my party.
These sites take a fair amount of time investment. To hit up against a fatal bug after the time has been put in, and/or after the social content has become public, is frustrating. And in an environment where reputation matters, I don’t need the Web site I’m using to make me look lame.
I urge these agile sites to think hard before releasing something without testing it. Because once it’s up, some of us Internet explorers will want to try it out. It’s hard to shed the last vestiges of belief that if it’s live, it must be working.
What seems to get lost in agile development is user research and testing. Additionally, the only people you can design for when you haven’t researched your audience are… people just like you. Two of my friends were just laid off from an agilely developed social content site that never had a vision or target audience and eventually withered away from disuse.
Obviously, some of the best innovations and disruptive technologies come from projects that were untethered by ‘what the user wants’ or can understand. But not every Web site is bringing something so amazing that it should claim this exemption. And, conversely, sometimes knowing the audience deeply can result in a highly refined design you wouldn’t get otherwise.
Either way, unripened fruit just tastes bad. In a way that makes one a little wistful of what could be.