Air Canada’s IT problems are almost as seasonal as Christmas itself.
Typically they’ve happened while I’ve been sleeping, and they’ve been resolved by the time I get into work. This is irritating because such incidents are about the closest we get to truly breaking news in the technology business. More dramatic than a virus outbreak, less opportunistic than a product release, Air Canada’s IT glitches also have a surprisingly limited shelf life. They happen, they’re solved, people grumble and the world moves on. More than almost any other Canadian enterprise, Air Canada challenges conventional IT wisdom. It has proven that downtime doesn’t really matter.
It’s entirely possible, for instance, that executives from Cisco, Q9, A&T or any of the other legions of firms who rail against the evils of downtime were among those waiting to board their Air Canada flight earlier today. They soon discovered that their boarding pass could not be printed. Instead, staff had to handle this process manually. Because humans are generally slower than their IT infrastructure (except, of course, when the IT infrastructure comes to a standstill), this caused some flights to be delayed around 30 minutes to an hour. That’s right: enough time to shop for a book or magazine, grab a coffee and donut or make another bathroom break. It was downtime, but it wasn’t really cataclysmic. It’s only the vendors and consultants who tell you even a minute of downtime will cripple your business.
This is not to suggest Air Canada won’t incur some costs because of the glitch it experienced Friday. But despite financial performance that has consistently lead to rumours of privatization and bankruptcy, Air Canada has managed to stay afloat despite problems much more serious than a computer glitch. Even the biggest problems associated with downtime – customer defection, for example – probably won’t apply. Air Canada customers with non-refundable tickets likely didn’t walk across the terminal to WestJet because their flight was pushed back.
The mainstream media focused on those most inconvenience by the downtime: a mother who had to re-pack her things and put her travel plans on hold, and an orthodox Jew who can’t take a cab or purchase anything during the period of limbo. These are exceptions to the general rule, however. Businesses like Air Canada not only experience downtime on a semi-regular basis, they have built in an expectation among their customers that it will happen. This could be more effective than all the backup and recovery systems money can buy.
No one would suggest companies should be cavalier about IT system reliability, of course, even if they don’t really lose much business or irrevocably ruin their reputations over it. Instead, maybe instead of swallowing vendor hyperbole IT managers should go through a risk assessment exercise that estimates what the real impact of downtime would be in their particular situation, and tailor these according to the key business processes they conduct. In Air Canada’s case, that would include not simply boarding passes but every step from taking ticket orders to serving them their in-flight meal. Not all of these processes would be IT-enabled, necessarily, but the ability to do them would obviously be hampered if IT system downtime blocked access to information or prevented a process that preceded it, creating a domino effect of sorts.
This may sound exhaustive, and I’m sure for most enterprise it probably would be, but like total cost of ownership, total cost of downtime would bring us one step closer to real business continuity. Now that its planes are flying again, Air Canada has a few moments to consider taking such a step. That’s one of luxuries uptime affords us.