An Engineering Organization’s Playbook for a Pandemic
Two weeks ago, I described how Thoughtworks was responding to the early stages of the Covid-19 outbreak, thanks largely to our offices in China seeing the crisis first. This is an update on how those measures have evolved as the situation has shifted—rapidly, and with little patience for the usual careful information-gathering cycle.
Our priorities haven’t changed. In order, they are:
- Protect the health and safety of our people.
- Help our clients, since that is our business.
- Slow community spread.
- Let regional teams drive responses based on local conditions, with global support and plenty of experience-sharing between them.
On that third point: most ThoughtWorkers are young and healthy, so infection is likely to be mild for them personally. But everyone who gets the virus can pass it on. Reducing travel and contact won’t necessarily reduce the total number of cases, but it can flatten the curve enough to keep health-care infrastructure from being overwhelmed, which materially improves outcomes for those who do get seriously ill.
The unknowns are still significant. We don’t have solid data on transmission mechanics or true mortality rates. It looks more dangerous than seasonal flu, possibly by an order of magnitude, while almost certainly not approaching the Spanish Flu. And if transmission drops with warmer weather in the Northern Hemisphere, there’s a real risk people will conclude the threat was overblown, skip prudent preparations, and face a worse outbreak next winter.
What Life Looks Like in Our Chinese Offices
New infection rates in China have dropped fast enough that we’ve reopened our offices there—but with significant changes. Government rules require an adequate mask supply before an office may open, and everyone is expected to wear a surgical mask in public. Everyone entering the office gets a temperature check. All surfaces are cleaned three times a day, seats are spaced at least a meter apart, and occupancy is capped at 50% of usual capacity.
There are obvious limitations to these measures. Temperature checks miss infectious people who don’t have a fever—the point is to encourage people who feel ill to stay home. Mask-wearing norms also vary by culture: common in Japan for people with colds, less so in Europe or the Americas, where masks are rarely seen on the street even now.
Global Measures and the Remote-Work Squeeze
Each country is following its own health authority’s guidance, but some policies are global. All internal travel and events are cancelled by default through the end of June. That includes sessions of Thoughtworks University, which we’d already moved from Xi’an to Brazil—they’ll now run remotely. The meeting that produces the Thoughtworks Radar is also going remote. These are events we’ve always believed needed face-to-face interaction, and we still do—but circumstances require pushing the boundaries of what can be done remotely. We’re also recommending that anyone skip events with 150 or more attendees unless attendance is truly necessary.
Many clients have barred external visitors entirely. That’s understandable—being infected by a colleague is one thing; being infected by someone from another company is another, particularly in a litigious culture like the U.S.
Remote work is the order of the day, but not everyone is set up for it. We’re working with clients who aren’t prepared, sharing our own experience with distributed teams. There’s a risk here: involuntary remote work is often set up badly, and a poor setup will leave a worse impression than remote work run well. Inceptions—the highly collaborative kickoff meetings for major work—are a particular challenge. They work remotely when the team already knows each other, as with the next phase of an existing product. But when people are strangers, a remote inception is much harder.
For teams that still need face-to-face collaboration, one approach is partitioning: split the team into two subgroups that never physically mix, with only half in the office at once. If an infection hits the team, only half its members are exposed.
Practical Frictions
Home setups are unequal. Some people have borrowed monitors and office equipment, but we can’t fix a lack of decent workspace. School closures add childcare duties and strain local telecom networks—home workers, remote schoolwork, and bandwidth-heavy games all compete for the same last mile.
We continue to require anyone who has traveled to a high-risk area, feels sick, or is high-risk themselves (or lives with someone high-risk) to work from home.
Because Covid-19 poses a greater danger to older people—and senior leaders tend to skew older—we’re planning for key people to fall ill. Every team should already have a plan for a member becoming unavailable suddenly; this crisis reinforces that principle. The same logic applies at the organizational level: identify who can step in quickly if a critical person goes down.
Scenario Planning and Organizational Resilience
Forecasting the economic consequences is even harder than forecasting the disease. Some companies will lose short-term revenue and bounce back; others will see it deferred until the economy recovers. Firms living close to the financial line may not survive the recession, even a short one. The most vulnerable are people in insecure jobs, often without sick pay, who can barely afford time off.
We’ve built a simple model to assess risk to our business. It’s simplistic, but it frames our thinking. We sketched several pandemic scenarios to guide early responses and to identify signals that indicate which scenario is actually playing out. So far, the biggest concern is demand: clients cutting back on new work as revenues suffer. We’ll keep refining the model over coming weeks and months. Being in China early was a disadvantage—we were affected before most—but it’s now an advantage, as we can learn from direct experience there.
A crisis stresses an organization’s policies and systems. Holding too firmly to normal controls limits the effectiveness of a novel response; letting them all go exposes the organization to large risks at exactly the moment its capacity to absorb them is weak. We balance these forces by pushing decision-making out to local teams and relying on networks between teams rather than a singular hierarchy for coordination. The trust and communication structures built during good times are an investment that pays off in difficult ones.
As I write this, the world seems rather scary. More countries are locking down, travel bans are spreading, and appeals for social distancing are everywhere. I’m actually relieved—it seems the world is finally taking the virus seriously. It’s important not to overreact; this is no Spanish Flu or Black Death. But without a bit of overreaction, we’ll see more death and suffering than we should. I’m encouraged by the seriousness my colleagues are applying to all of this. Facing dramatic, unpredictable events with calm and compassion is a core strength of this organization, and I hope the rest of society takes a similar attitude.
