Welcome to Marketing BS, where I share a weekly article dismantling a little piece of the Marketing-Industrial Complex — and sometimes I offer simple ideas that actually work.
If you enjoy this article, I invite you to subscribe to Marketing BS — the weekly newsletters feature bonus content, including follow-ups from the previous week, commentary on topical marketing news, and information about unlisted career opportunities.
Thanks for reading and keep it simple,
How fast is COVID-19 spreading?
Last week, researchers from the Los Alamos National Laboratory released a report that analyzed the spread of COVID-19 in China. An April 8 article in Bloomberg highlighted the report’s conclusions:
The new coronavirus raced through China much faster than previously thought, a U.S. research team said, suggesting that extremely widespread vaccination or immunity will be necessary to end the pandemic.
Each person infected early in the epidemic in Wuhan probably passed the virus to an average of 5.7 other people, according to a mathematical analysis from Los Alamos National Laboratory. That’s more than twice what the World Health Organization and other public health authorities reported in February.
The team’s results are specific to the Chinese outbreak. If they hold true elsewhere in the world, the pandemic may be more difficult to control than some authorities had modeled.
To evaluate the speed at which a virus infects a community, researchers consider two key metrics:
Doubling time: the amount of days it takes for the number of positive cases to double. The shorter the doubling time, the faster the virus will spread.
R0 (pronounced “R-naught” and often referred to as the “basic reproduction number”): the number of individuals directly infected by one infectious person (in a completely susceptible population). With an R0 less than 1.0, a disease will eventually fade out of existence. If R0 is greater than 1.0, a disease could — in theory — infect everyone in the world.
In late March, the WHO estimated COVID’s basic reproduction number as 2.2 — far lower than the R0 of 5.7 cited by the Los Alamos National Laboratory.
Information is Beautiful created an interactive chart that compares the contagiousness of COVID-19 to other diseases, including the following ones:
2.2 — Ebola
2.2 — 1918 Spanish Flu
2.2 — COVID-19 (WHO estimate from March)
3.5 — Polio
5.0 — Smallpox
5.7 — COVID-19 (US estimate from April)
8.5 — Chicken Pox
9.0 — Measles
Some important context: an R0 of 2.2 is still really, really dangerous. In that scenario, the virus only needs 30 “cycles” to spread from a single infected person to everyone on the planet. But an R0 of 5.7 would spread even faster, reaching everyone in the world within just 14 “cycles” — that’s the power of compound growth (a topic I covered in a recent post).
To be clear, no one is suggesting that — even with an R0 of 5.7 — COVID-19 will infect every single person on the planet.
After people recover from COVID-19, they develop an immunity to future infections (most medical researchers believe this concept to be true, based on similar viruses). As such, with each COVID-immune person in a community, there are fewer people for a COVID-positive person to infect. Over time, the R0 continues to decrease; if R0 drops below 1.0, then the virus would begin to dwindle.
If you’ve been following COVID-19 in the news, you’ve probably heard the term “herd immunity” — especially in the context of the UK’s aborted plan to fight the pandemic.
As defined by Science Alert:
Herd immunity is an epidemiological concept that describes the state where a population is sufficiently immune to a disease that the infection will not spread within that group. In other words, enough people can't get the disease — either through vaccination or [previous infections] — that the people who are vulnerable are protected.
For an R0 of 2.2, herd immunity would kick in when approximately 55% of the population develops immunity.
If, however, the latest estimate for COVID-19’s R0 (5.7) is correct, herd immunity would not be achieved until approximately 82% of the population develops immunity. In practical terms, a level of 82% could only be reached with mass vaccinations — not from just letting the disease “run its course.”
Bringing R0 down
If you think back to high school science classes, you might remember some mathematical details about the force of gravity (objects fall at 9.81 meters per second squared). Basic reproduction numbers, in contrast, are not fixed at specific levels; various factors can increase or decrease an R0.
Changes in behavior — like social distancing and stay-at-home orders — will lower the R0 of COVID-19, hopefully below the 1.0 threshold that causes diseases to fade away. Beyond lockdowns, other factors could impact R0. Think about different styles of cultural greetings. I expect that, all things being equal, cultures where people greet one another with a bow would have a lower R0 than handshake cultures. And handshake cultures, all things being equal, would have a lower R0 than ones that embrace and kiss cheeks.
Generally, R0 refers to an entire population, but you can also evaluate individual people’s R0. Some people might contract COVID-19 and not infect any other people; their R0 is zero. But other infected people might spread COVID-19 to 100 or more people; people with R0>100 are sometimes referred to as “super spreaders.” In South Korea, authorities tracked all of the people with close contact to COVID-positive individuals. One woman — who attended multiple church services and a buffet lunch — interacted with 1160 people before she tested positive for the virus.
Loyalty and Flight
The idea of COVID “super spreaders” mirrors the way that a lot of marketers think about their customers — especially the “super-loyal” ones that proselytize for their brand.
Many consultants love to push loyalty programs. With fancy charts and slick presentations, consultants explain that loyal customers spend way more money than non-loyal customers. As an added bonus, the loyal customers stick around a lot longer. The consultants’ enthusiastic advice usually sounds something like this: “If you convert all of your non-loyal customers to loyal ones, your business would grow by an order of magnitude!”
This strategy hinges on the belief that by changing your messaging, you CAN improve the loyalty of your customers. In reality, though, this concept only works on the margin, with (generally) minimal impact.
The better approach? Find new customers who will quickly develop loyalty to your brand.
Let’s imagine a business that only attracts two types of customers:
Loyal customers, with a churn rate of 1% per month.
Flighty customers, with a churn rate of 50% per month.
For this exercise, let’s suppose that you attract 100 NEW customers, split equally between the Loyal and Flighty types.
What is the churn rate for your business?
Start by adding the churn rates for the two categories — 50 and 1 — and then dividing by two. Expressed as a formula, that works out to ((50+1)/2 = 25.5). Hence, this hypothetical company’s churn rate is 25.5%.
BUT, the 25.5% figure only applies to the first month. For subsequent months, the composition of customers will change significantly: half of the Flighty customers keep departing, but only 1% of your Loyal customers disappear each month.
The following chart illustrates the rapid change in your customer base:
At the start, your churn rate is 25.5% (the simple average of the two customer groups). After one month, half of the worst customers drop off, which lowers the churn rate to 17.4%. By the sixth and final month, the churn rate is way down to 1.8%.
Of the 50 Loyal customers, you retained 47.1 of them. But almost all of the Flighty customers flew the coop over the previous six months — only 0.8 Flighty customers stuck around.
Many consultants and marketers, though, try to convince companies that new strategies can — and will — improve the loyalty shown by customers. “If we can find ways to hold new customers for the first three months, then they’ll be loyal customers for life!”
Their analysis clings to the (false) assumption that companies can somehow “trick” the Flighty customers to stick around for the first three months. Consultants point to the type of information shown in the previous chart and say things like, “By month 3, your churn rate has dropped to 6.6% — a huge improvement over the 25.5% churn rate at the beginning.” But of course the churn rate dropped — many of the Flighty customers departed, so your customer base now consists almost entirely of Loyal customers! You haven’t changed the churn rate of your customers, you’ve simple changed which customers are left.
What happens next (in most cases)? The new marketing strategies fail to “promote” your non-loyal customers to loyal ones. In the board room, the results disappoint everyone, and people brainstorm feeble excuses to explain the let-down.
Changing a customer’s behavior is HARD. Enticing a customer to stick around for one extra month does not fundamentally change their overall spending habits, their particular tastes, or — in the broadest sense — who they are as a person. In some instances, you CAN influence customer behavior in the short term, but that rarely changes their underlying nature. Fact: some customers are a better fit for your product than other people.
If changing behavior is hard, then what is the best way to attract more loyal customers? Answer: just get more customers period. Some of your new customers will quickly identify an ideal fit with your product; these people will develop loyalty and continue to churn at a low rate.
Marketers refer to this concept as “customer heterogeneity.” In simplest terms: customers are all different. The metrics of your business are largely driven by customer differences, NOT the business strategies you implement to alter customer behavior.
Let’s circle back to R0 values.
When COVID-19 strikes a community, who is more likely to be infected first: (1) an introvert who enjoys spending nights at home, or (2) an extrovert who loves social outings? Obviously, you would expect the extrovert to contract the virus first.
Now, suppose both of these people are infected. Which one would be more likely to spread the virus to other people? Once again: the extrovert. In terms of R0, I think it’s reasonable to believe that extroverts have a higher R0 (maybe 10.0) and introverts have a lower one (maybe 0.2).
When COVID-19 first enters a community, I think the virus might disproportionately infect extroverts. If you measured the R0 of this hypothetical community as a whole, you would find something close to 10.0 (because at the early stages, almost all of the cases are extroverts).
What happens next? Over time, the initially infected extroverts will (hopefully) recover from the virus, reducing their ability to (1) spread the virus to any more people, and (2) contract the virus again. Of course, some introverts would contract COVID as well.
At some point, the positive cases in the community might be split 50/50 between the extroverts and introverts. In that case, you can determine the R0 for their entire community by using a blended average of the two groups — just like we saw with the discussion of customer churn in the previous section. The math: R0 is ((10.0+0.2)/2 = 5.1).
In this example, the community’s R0 has dropped from 10.0 in the early onset to almost half (5.1) after a period of community transmission. Notice that the R0 dropped not due to any specific intervention, but because of “customer heterogeneity.”
How impactful is customer heterogeneity?
If every person behaved identically, then an analysis using customer heterogeneity is probably irrelevant. But we know that people do NOT behave in identical ways. Think about the people you know and the massive differences in their personality traits, habits, etc.
For the spread of COVID-19, I expect that variance in R0 is a very important (and overlooked) factor. Recall that the WHO estimated COVID’s basic reproduction number as 2.2, a much lower R0 than the 5.7 reported by the Los Alamos National Laboratory. Regardless of whether COVID-19’s real R0 is closer to 2.2 or 5.7, I suspect there could be a wide variation in the basic reproduction number of individual people; some people could have an R0 lower than the 2.2–5.7 range, while others (the “super spreaders”) could have a much, much higher R0. In other words, the R0 identified by researchers might have been skewed by a small number of people with an astonishing capacity to spread the virus.
Media outlets have published myriad articles about super spreaders (Spring Breakers in Florida, Smurfs in France, and a particularly chilling story about a choir north of Seattle). I have also found some more analytical perspectives about variation in R0 (a case study from Austria and a general consideration of power-law distribution).
But most of these stories focus on circumstances, like a shared social event or a specific occupational group. I recognize the obvious importance of circumstances; for instance, we know that frontline workers are at much higher risks of exposure. That said, I think more analysis of “customer characteristics” — like introversion versus extroversion — could inform a greater understanding of how community transmission actually occurs.
In the corporate world, we often learn that measurements of churn rate under-appreciate the impact of intrinsic human differences and overvalue the impact of strategies intended to “convert” people into loyal customers.
With COVID-19, perhaps we are falling into the same logic trap, attributing too much credit to specific actions (lockdowns, etc.), but not enough attention to customer heterogeneity. That theory could explain why coronavirus outbreaks have spread at radically different speeds across some regions compared to others. Maybe those regions have varying distributions of “customers” for the virus, just like different marketing channels will deliver distributions of customers with dissimilar “base” churn rates.
Of course, I am a marketer and business strategist, not a virologist. Please take everything I say with a grain of salt.
Keep it simple and stay safe,
If you enjoyed this article, I invite you to subscribe to Marketing BS — the weekly newsletters feature bonus content, including follow-ups from the previous week, commentary on topical marketing news, and information about unlisted career opportunities.
Edward Nevraumont is a Senior Advisor with Warburg Pincus. The former CMO of General Assembly and A Place for Mom, Edward previously worked at Expedia and McKinsey & Company. For more information, including details about his latest book, check out Marketing BS.