Listeners to the most recent episode of Digital Campus will know that I’m a fairly heavy user of Twitter, the weirdly addictive and hard-to-describe microblogging and messaging service. But anyone who uses the wildly popular service regularly will also know that the company’s service architecture has not scaled very well. During the last month or so, as hundreds of thousands have signed up and started “tweeting,” it has sometimes seemed like Twitter is down as often as it’s up.
Considering the volume and complexity of the information they’re serving, and the somewhat unexpectedness of the service’s popularity, I tend not to blame Twitter for its downtime. As a member of an organization that runs its own servers (with nowhere near the load of Twitter, mind you), I sympathize with Twitter’s situation. Keeping a server up is a relentless, frustrating, unpredictable, and scary task. Yet as a user of Twitter, I still get pretty annoyed when I can’t access my friends’ tweets or when one of mine disappears into the ether.
It’s clear, however, that Twitter is working very hard to rewrite its software and improve its network infrastructure. How do I know this? First, it seems like some of the problems are getting better. Second, and more important, for the last week or so, Twitter has been blogging its efforts. The Twitter main page now includes a prominent link to the Twitter Status blog, where managers and engineers post at least daily updates about the work they’re doing and the problems they’re facing. The blog also includes links to uptime statistics, developer forums, and other information sharing channels. Twitter’s main corporate blog, moreover, contains longer posts about these same issues, as well as notes on other uncomfortable matters such as users’ concerns about privacy under Twitter’s terms of service.
Often, an organization facing troubles—particularly troubles of its own making—does everything it can to hide the problem, its cause, and its efforts to fix it. Twitter has decided on a different course. Twitter seems to have realized that its very committed, very invested user base would prefer honesty and openness to obfuscation and spin. By definition, Twitter users are people who have put themselves out there on the web. Twitter’s managers and engineers have realized that those users expect nothing less of the company itself.
As a Twitter user, the company’s openness about its difficulties has made me more patient, more willing to forgive them an occasional outage or slowdown. There is a lesson in this for digital and public historians. Our audiences are similarly committed. We work very hard to make sure they feel like we’re all in this together. We should remember this when we have problems, such as our own network outages (CHNM is experiencing one right now, btw) and technical shortcomings.
We are open with our successes. We should be open with our problems as well. Our audiences and partners will reward us with their continued loyalty and (who knows?) maybe even help.
