Last Friday, as the sun rose, the internet went dark for millions of Canadians. Rogers’ wireless and wired networks were both down, allegedly following a botched software update overnight.
The outage lasted all day. The disruption was widespread, taking workplaces, emergency services and banks with it. The magnitude was large enough that I expected statisticians to see it in the GDP statistics. For the rest of us, what won’t show is the shock of being so unexpectedly disconnected in our new hyper-connected world.
While there have been the usual calls for an investigation and concerns about the lack of diversity among our ISPs, the call we need to answer is a wake-up call. Between the two Rogers outages in April 2021 and last week, we know the systems are technically more fragile than believed. Add to that concerns about potential cyberattacks, and would-be attackers are now aware of that too. There is an urgent need to manage these risks.
The path we have chosen so far is what economists call protection. We relied on ISPs to manage their own resilience and reduce the likelihood of bad events happening. While this may go a long way, there is no sure way to prevent disruption.
What we need insurance to complement protection efforts. The question is: if there is a disruption, how quickly can we mitigate the costs? The current approach is to rely on the networks themselves to scramble the resources and deal with the outage. But last week it took a day, involved no clear timeline and left some areas without coverage for days.
There are alternatives to a centralized solution. Some companies maintain links with more than one supplier. They can use Rogers as their primary wireless service provider, but if that network goes down, they switch to Bell or Telus by swapping a few cables. Or for cell phones, they can keep a spare SIM card handy. Duplication in this manner is expensive and only the most critical services are likely to pay a duplication premium as insurance.
Could this type of solution be widely available? The technical costs are probably low. Most Canadian establishments have more than one Internet provider providing the infrastructure at their doorsteps. For phones, the federal government is already pushing to improve the network’s roaming options following the Rogers outage.
But the solutions can be consumer-centric. For example, smartphones are increasingly being built with e-SIMs which are built into the hardware and therefore do not require physical switching. In other words, the actual costs of duplication have already been spent.
What prevents them from being an insurance policy for people is a contractual matter. Multiple points should be readily available for use at all times. You should be able to trade your Rogers connection for Bell, Telus or others on your own terms and have to pay (or charge your current network) for it after the fact. This will allow people to get back online quickly without having to wait for a centralized solution or a service repairman to come to their doorstep. In principle, breakdowns will become a minor inconvenience.
An insurance policy would also be beneficial for the networks. Rogers had two problems last week. First, the network had to be brought back online. Second, he had to somehow communicate with his customers to find out what was going on. For me, it forced me to use the only radio I have left these days in my car!
From Rogers’ side, it must have been overwhelming and also caused them to sew things up in a rush rather than a thoughtful way. Having an insurance option would give them leeway.
Nonetheless, I suspect that developing and implementing a contractual framework that enables customer-level assurance will require government coordination and perhaps pressure. We will need to ensure that the “prices” of insurance are reasonable and even consider how this might reduce switching costs for consumers and stimulate competition between networks.
Governments still have a heavy hand on insurance. We should expect them to take leadership on critical infrastructure as well.