I wrote this back in March for a client who shared it on their internal network. Now that we’re hitting the predicted Fall/Winter spike, I thought it would be interesting to revisit it.
The first draft (February 2020) was really really short:
The second one (March 2020) was a bit longer:
I’ve been watching the progress of this new coronavirus for a couple months, and I have to admit I initially didn’t think it was that big a deal. My previous experience with pandemics told me that it was unlikely this would spread into Europe, much less across the Atlantic, and that media reports tend to overstate and sensationalize the danger…but then at some point in February I had a moment of clarity and realized that I was being guilty of the same cognitive behavior that so frustrates me in other people.
“Wait a second; this is an emerging exponential event that requires immediate action! I literally do this for a living!”Me, in March
It was the emotional fluctuation that clued me in. It felt really familiar. Like so many others, I’d been switching between “I mean, is this really serious though?” and “Why hasn’t literally everything been shut down except hospitals?!?!” The frantic voice, beating on the walls yelling for someone to LISTEN and ACT, combined with the soothing voice that tries to reassure, advising to stay the current course, everything’s fine…these are the exact companions that innovators, especially those working inside an established corporation, live with constantly. We have uncomfortable knowledge about what’s happening and what’s coming. We try to get decision-makers to understand it and act on it. Overcoming organizational inertia is crazy hard, and we’re plagued with self-doubt…even when we know we’re right, we know we are risking our career and social standing in sounding the alarm or suggesting a new way.
Decision-makers prefer “safe” partial actions, baby steps that hew as closely as possible to the status quo, and that look like a response without actually addressing the core issue. In a large corporation, linear, short-term, direct cause-and-effect actions are the only ones that can make it through the multitude of governance boards and committee meetings. That type of action only addresses the tip of the iceberg…which means the organization gets hit by the other 90% of the ice they didn’t see.
Decision-makers don’t want to take large-scale actions that will have significant impact on, well, everything. Even if they suspect that might be the only option, they try to slow-roll it, talking it out, “socializing” it in PowerPoint decks, making sure everyone’s “bought in” and everyone’s “equities are considered,” when in fact a large, systems-level response should be started right away to ensure that the organization has the time it needs to pivot.
This happens in part because leaders and decision-makers feel that they must come up with the perfect plan right away, rather than being comfortable setting a goal and stepping out, then being willing to pivot as they get new information. They don’t tend to delegate responsibility for solving parts of the problem.
Am I talking about asking an organization to make a significant change so that it remains viable in the market? Or asking a nation to make a significant change to mitigate the damage from a pandemic?
The Umbrella Academy
A few years ago, a friend of mine who is a cognitive scientist gave me a great metaphor for the challenges in convincing someone to adopt or accept innovation:
There’s a guy whose house is exposed to the sun all day. He’s got 100 umbrellas and he spends his day readjusting them, constantly shifting the angle of each one to ensure that it’s between the sun and his house, keeping the house in shade. He can’t enjoy the house or relax, but hey, at least he’s got a solution. A stranger approaches; she’s got an acorn and offers to help the man. If he will give her one of his umbrellas, she will turn it upside down, fill it with good soil, and plant the acorn. With a little water, in a few years a small tree will shade some of the house. In a few more years, a beautiful oak will shade most or all of the house, and the man won’t have to spend his days running in circles trying to adjust umbrellas.
What do you think the man does?
He tells her no thanks. He can’t give up even one umbrella, because that will throw off his groove. He’ll have a gap. Also, maybe he’s concerned that he’ll the guy who manages 99 umbrellas, which isn’t as good as being the guy with 100 umbrellas. He doesn’t doubt that her solution will work, but then he won’t have a reason to exist anymore. And that’s all the time he can give her because he has to get back to adjusting umbrellas.
He cares about the process, not the outcome, and the press of process leaves him no time to really think through the opportunity and risk.
I think this is a perfect parable for why it’s so hard to carve out space, time, and resources to create a new capability or business line in a corporation, even when the change would be beneficial. It also evokes the saying “the best time to plant a tree was 20 years ago.”
I also think it works as a parable for why US institutions have been slow to commit to closing for the coronavirus (and why the very innovative SXSW was one of the first to do so): the decision-makers can’t let go of the status quo. I get it – I mean, as an innovator I do not live in the status quo, I live in constant extreme uncertainty. And even I was slow to acknowledge what’s happening.
Decision-makers also have multiple systemic incentives working on them, and it’s often not clear when the weight/value/importance of those change. For instance, if you send everyone home to work remotely, do you really need all that mid-level management? There will suddenly be a lot fewer meetings, which means you won’t need to hire employees to do the work that the person who goes to all the meetings isn’t doing. How do you weigh not wanting to fire people against wanting to have a big company against the prediction of a serious pandemic?
There’s also a friction between those who are motivated by the outcome (Safi Bahcall calls this “stake” in Loonshots) and those who are motivated by their social or financial status (“rank”). If you are motivated by your rank, you won’t quickly move to change things in a way that might undercut that. If you are motivated to achieve a particular outcome, you’ll try to move quickly toward it. Thus an innovator might say “we’ve invented a digital camera; we should get those to market!” while a corporate might respond “I am the executive of a film company, not a gadget company” (and you get Kodak). A doctor might say “we need to enact a full quarantine to stop the coronavirus,” but a CEO might respond “but that will cut into my profits” and a mayor might respond “that will put people out of work and make my city scary.”
Comparing Apples to Buicks
This leads into the next reason it’s hard to make space in an “execute repeating tasks” organization for new stuff: perspective and comparisons. This is a classic cognitive bias that everyone falls into: are you comparing the right things? (If you are interested in this sort of topic, I recommend basically everything by Tim Harford)
When I first started hearing about Covid-19, I initially compared it to previous pandemics I was familiar with…but when it caught hold in Iran and Italy I started making poor comparisons. You may have done or seen the same: “200 people died in Italy in 24 hours and there are calls to quarantine, but 157 people die the US from car crashes alone and we aren’t restricting car travel!” I hope that you can see why this makes no sense as a comparison; I can think of a few reasons.
It turns out that just because you can compare two things doesn’t mean you should; even if they seem like they are both about national death rates from particular sources, you aren’t going to get useful and actionable insights.
Most of the comparisons I’ve seen in the media, as we try to find a good frame of reference for Covid-19, are with the seasonal flu. Statnews makes the case that the two disease sets have very different trajectories through a population. When you are a leader trying to determine whether to stay open and deliver value to your customers, you need to know whether the cost will be some employee sick days and a very small chance that someone dies because your office was a vector, or will be dramatically overstressing the health care system and contributing to thousands of deaths.
Of course, we know from evidence what the seasonal flu risks are; we don’t know what Covid-19 risks are. We have data from China (questionable), South Korea (better), and Italy (better). None of those are great comps for US geography, density, and population distribution. South Korea has advanced health care; Italy has the second-oldest population in the world, and almost all of them are smokers. So we are largely guessing.
Using Models to Make Better Guesses
The flu is a coronavirus, so while none of us are immunologically prepared for SARS-CoV2, we can probably still start with flu models. The CDC has these for each year, and they are all very similar.
This is a classic exponential curve, and why I started taking Covid-19 seriously. I spend my days convincing people who live in the linear growth world that a particular trend is actually exponential. The reason this is tough is that an exponential curve looks like it’s growing less than a linear increase at first, then it explodes. If you trace a straight line from the start of the chart above to the peak, you’ll see that for the first ten or so reports, the growth is lower than linear. That can introduce a false sense of security. (This isn’t a great example because the average linear growth that a business strives for will take MUCH longer to get to the same high point than exponential growth will, but still).
Managers love linear growth. It’s steady and predictable and good for quarterly meetings. OTOH, managers are responsible for delivering ROI quarterly and hate hearing “we’re below expectations now, but any day we’ll hit the “hockey stick” and zoom up!” Most people hate hearing an innovator or Cassandra say “I know it doesn’t look like much now, but any minute now we’re going to be up to our ears in Tribbles!”
There are several epidemiological studies that predict a doubling of coronavirus cases every 6 days. (Statnews points out that we can expect small variances: “Confirmed cases may appear to rise faster (or slower) in the short term as diagnostic capabilities are ramped up (or not), but this is how fast we can expect actual new cases to rise in the absence of substantial mitigation measures.”)
Assuming no intervention or mutation, that means the US is looking at:
- May 1: 1 million cases
- May 7: 2 million cases
- May 13: 4 million cases
- May 19: 8 million cases
Assuming the death rate likewise remains at 2.5% of cases, that means:
- May 1: 25,000 deaths in the US
- May 7: 50,000 deaths
- May 13: 100,000 deaths
- May 19: 200,000 deaths
That’s not the same trajectory as traffic fatalities AT ALL. (And we did put a lot of mitigations in place to reduce traffic fatalities over the past decade.)
Except it’s NOT Exponential
Just as the annual number of flu cases doesn’t increase exponentially forever, the number of Covid-19 cases won’t either. Most people with Covid-19 recover and it appears they stop being transmitters after having it. People who haven’t gotten it start taking protective measures; all the stuff that is being advised to help “flatten the curve.” This causes the trend line to peak and then descend.
But then, just as the next season of flu comes around, we are likely to see another peak of Covid-19 cases after a few months, as the virus mutates, or people stop self-quarantining. The chart looks more like a sine wave than a one-way hockey-stick curve.
This peak-and-trough means that it’s easy to underestimate the risk early on and then overestimate the risk as it falls, then feel like it’s solved and not prepare for a return, be surprised, underestimate the growth, rinse and repeat.
We will need to watch Covid-19 very carefully to not get surprised by a second (or third) peak. I’m beginning to wonder if the unequal distribution of Covid-19 risk (the elderly are MUCH more likely to die from it than the young) will muddy the data.
Another thing that might muddy the data is if America embraces “flatten the curve and raise the line.” Obviously flattening the curve by taking self-protective measures would, you know, flatten the curve. If we institute a nationwide lockdown that will flatten it further. If we then invest in increasing our health care capacity as China and South Korea have done (raising the line), we’ll have less impetus to stick to a hard curve-flattening and can expect to see some wobble. However, overall the exponential rise should be disarmed and the peaks lower, which means fewer cases and fewer deaths.
Back to innovation
If you run in innovation circles you hear a lot about “failing forward” and “good failure.” These are true and good concepts, but situational: they refer to, say, putting a rough prototype in the hands of early users, learning all the things they hate about it, and actually improving those so the next version is better. It’s analogous to practicing a sport or a musical instrument, noting your mistakes, doing the work to reduce or remove them, and practicing some more.
But there is bad, even catastrophic, failure in innovation. What if a performer or athlete just doesn’t practice? What if they’re like “Yeah I think I have a concert coming up but I’m not even going to check to see if my violin needs strings?” If we just totally fail to see what is happening and act accordingly, we’ve failed the bad way. If we stop introducing new capabilities and just let our products or services become obsolete, we’ve failed.
We have to simultaneously maintain a strategic mindset (what is our value prop? what is our guiding star?), an operational mindset (what are the main ways to get there?), and a tactical mindset (what do we need to do RIGHT NOW?). I’ve written about the cynefin framework before and it’s always helpful for this. It may also be helpful for decision-makers who are trying to figure out how to respond to Covid-19 right now.
We’re in the Chaos quadrant. That means OODA: Observe, Orient, Decide, Act. Do it fast. Do what you need to do to get off the X and move to Complex or Complicated. I personally think the Covid-19 situation has enough moving parts that it’s going to be Complex at best, but maybe smarter people than me can shift us into just Complicated. From there, everything is figureoutable.
But right now, we are in the ramp-up to exponential, and the best time to act has already passed.
The third draft…is this one (Nov 2020).
I kind of hit and kind of missed. Geography (coastal vs. Midwest, urban vs. rural) has played at least as big a role. The exponential peak wasn’t as bad, but the rolling peaks are still bad. We are now hitting the second big peak with no real change from March.
I’d love to hear your thoughts on this…how has your thinking about “The Rona” (as it’s called now) changed? What have we learned and what do we still need to learn? Will we pivot in time to reduce the third peak?