The COVID-19 Persona Analysis: A Profile of The Neighborhoods Infected With and Recently Unemployed By The Coronavirus

6 May 2020 -- First, some good news. COVID death toll in the US continues to decline. COVID is no longer the #1 leading cause of death in the US (see chart below). The rate of confirmed cases is decreasing in nearly every state. Overall, the weekly rate of case growth has finally dropped below the 20% -- it stands at 19% as of yesterday. (Minnesota, Nebraska, and Iowa are outliers with high week-over-week growth rates).

COVID-19 Deaths vs Other Common Causes

The first tragedy of COVID is the impact on lives -- more than 250,000 deaths globally and more than 70,000 in the US. The Second tragedy is the impact on the livelihoods. A global figure of unemployed is difficult to estimate. In the US, the figure stands at over 30 million recently unemployed in a mere six weeks. The scale dwarfs anything on record - but we can and will rebuild. Part of that process is understanding more precisely who COVID affects.

COVID-19 New Unemployment Filings

What I aim to do with this article is to show you the people behind the trend lines. Marketers use personas to put a face to data. Let me do that for you with two aspects of COVID-19. First, those who have tested positive. Second, those that are recently unemployed.

If we want to understand COVID-19 health and economic implications, we have to better understand who it is disproportionately affecting. A tool I have used over the years in persona analysis takes thousands of data points about where people live, work, shop, their attitudes, intentions, behaviors, demographics and hobbies and then looks for commonalities by zipcode. Companies like Experian create segments and composite profiles based on this data. Analytics people like me can match the case counts by zipcode and find the segments that over index. Each segment is about 1 to 2% of the US households, which gives a fairly specific profile. It doesn't replace a true first person account, but in some ways it is better because it is a representative view of all those we are analyzing.

I crunched the COVID case data by zipcode for the state of Illinois, the two largest counties in Nevada, and the cities of New York and San Francisco. I also ran the data for the state of Texas on the recently 1 million unemployed. I would have done a country wide analysis, but the US does not have any data source for COVID by zipcode that I could find. Certain states and counties publish, but many don't -- or they provide a map with no ability to analyze the data further. (If you want to analyze your city, county or state and have the case count or unemployment by zipcode, I'd be happy to help you.) Finally, I used Experian MOSAIC, a popular global solution for this type of analysis. Experian periodically adjusts the names, and if you are power user, forgive me for mixing some of the new and older names for the segments as I crunched the data.

Persona Analysis Characteristics Used For Segmentation

The first part of the analysis is a portrait of the households that were more likely to test positive. The second analysis are the households most likely to have become unemployed due to COVID.

Personally, when I review persona data, I suspend any judgment and think of someone I know and love that falls into this segment. I try to imagine what their hopes, dreams, fears and aspirations are. As you get to know who is disproportionately affected by COVID, consider what it will take to help them recover from COVID. If you are a business owner or marketer, do a similar analysis of your customers so you know how they overlap with these segments. It will help you better understand your path to rebuilding your business.

Part of my broader effort in this analysis is to build a model of the economy in a post COVID-19 world, so we can best understand how to recover and rebuild. First step is understanding who has been affected by COVID. Second step is to figure out how our spending behaviors will change -- which will likely differ greatly by segment. Stay tuned for more data and analysis on how our economy and industries will change.


Meet the Top Three Segments Most Likely To Have COVID-19:

Meet Shawn & Wendy. They are members of a segment of mature, middle income, single adults and families living in urban areas. Experian calls them Mid-Scale Medley.

Persona Analysis Characteristics Used For Segmentation

According to Experian, Mid-Scale Medley are both singles and couples living in older, city neighborhoods that have been bypassed for gentrification. These working-class, blue-collar laborers often live in worn houses and funky apartments. Most are middle aged and a little over half are childless. Their educational profile is a mix of some high school, some with diplomas and many who have taken some college classes but never graduated. The majority work at lower-echelon or manual labor jobs in transportation, food services and construction.

Many have managed to buy their homes, which typically were built more than a half-century ago. Mid-Scale Medley lead unpretentious lifestyles and are happy to take advantage of nearby city amenities, and on weekends, maybe go on a camping trip. Because they work long hours, they don’t spend a lot of time at home but, when they do, they enjoy painting, needlework, listening to music and playing video games. Ever in search of opportunities to make extra money, they might buy a lottery ticket. Most prefer to shop at nearby stores, typically buying what they need at the moment and ignoring the designer fashion of high-end boutiques. They shop discount retailers like Family Dollar and Kmart. Friends and family often ask for their opinions on a range of products. With fewer than half owning cars, they rely on public transit to get to their jobs and downtown entertainment. They have a moderate tendency to travel domestically, taking vacations by plane, rental car and RV.

This segment is more than 2.2 times more likely to have COVID-19 compared to other communities in the area.

Meet Roger & Connie. The Sophisticated, down-scale singles and couples living in modest, exurban small towns. Experian used to call this group Small Town, Shallow Pockets (which is still reflected in my table) and now calls them Small Town Sophisticates.

Persona Analysis Characteristics Used For Segmentation

According to Experian: Small Town Sophisticates are older, unmarried empty nesters in second-tier cities and exurban towns. Their lifestyle is pure small-town America. Most residents are over 50 years old and include a mix of single, divorced and widowed individuals living in downscale neighborhoods. Less than 10 percent have a college degree, and the majority work in service-sector and blue-collar jobs. Nearly 15 percent are already retired.

Their neighborhoods, often found in cities and towns that have seen better days, are quietly aging. The housing stock is a mix of bungalows, cottages and ranch houses typically built in the first half of the 20th century. Most houses are small and their lots modest. Home values are only a third of the national average and yards are rarely landscaped. In these areas, status is a new truck out front. This segment is 2.1 times more likely to get COVID-19 compared to other communities in the area.

Meet Billy and Tonya.They are Working-class, middle-aged couples and singles living in rural homes. Experian calls them Touch of Tradition.

Persona Analysis Characteristics Used For Segmentation

According to Experian, Touch of Tradition live in small, isolated communities that are home to no more than a few thousand inhabitants spread across a rural landscape. Many towns are so small that they typically consist of little more than a church, campground and a general store that doubles as a cafe. These households tend to contain middle-aged couples and singles living in mostly compact houses and mobile homes. With their modest educations, most work at blue-collar and service jobs. Even though a small percentage works as farmers, the number is over four times the national average. The pay is low, but expenses are also modest, and these folks have crafted unpretentious lifestyles in their remote settings.

Overall, this segment is twice as likely to get COVID-19 (within Illinois, more than 4 times more likely).

Putting COVID-19 In Perspective

Like other zipcode based segmentations, Experian shows how the segments differ by Income, Age and Family Status. For example, those segments closest to "High Income" have the highest income. The red circles are the top ten segments with Covid-19. Notice how not a single one is above the half-way point on income?

Persona Analysis Characteristics Used For Segmentation

Since testing in the US has missed many people infected, I wondered if, perhaps, these segments were tested more and therefore they show up with more cases. The state of Illinois was good enough to provide both tested and confirmed -- so I could check. As it turns out, more affluent segments are more likely to be tested by almost a ratio of 2:1 versus the poorer segments. Therefore the concentration of COVID in the lower income is not due to more testing - instead it is in-spite of less testing. Therefore, the concentration in the poorer segments may be understated.

Regional Differences

Different regions have some differences in which segments live in the area, and which over index on getting COVID. Below is the table by region. The table is ranked from high income and education to lower income and education (Experian numbers segments from A01 to S71 - The A group being the most affluent). In terms of the Red/Green color coding of the data, red means lower chance of getting COVID while green is higher. The number in each cell is the comparison to average for the area -- 1.00 means the segment is equally likely to get COVID. 0.50 means the segments is half as likely. 2.00 means twice as likely to get COVID. This number is the index.

NYC is on the far right of the table. NYC's highest indexed segment is a group called Striving Forward (see persona below). SF, the next column to the left has a segment called Urban Survivors as the highest indexing segment with COVID. Illinois; similar to the overall average, has Touch of Tradition is a segment that is 4.12 times more likely to get COVID and Mid-Scale Medley is 3.9 times more likely than average to get COVID (both segments are introduced above). The next two columns to the left are counties in Nevada. Clark is in the south, where Las Vegas is located. It shares the Touch of Tradition as the number one segment more likely to get COVID. The second highest segment is Countrified Pragmatics, which I profile below. Finally, to the left most is Washoe County. It is a large county that includes Tahoe, Reno, and a large rural area stretching to the Oregon border. Washoe is the only area I analyzed where an affluent area breaks the 2.00 more likely index. This is due to an outbreak in Crystal Bay, on Lake Tahoe. However, most of the outbreaks are over a hundred miles from the lake in the poorest communities of Washoe. The segment Small Towns and Shallow Pockets is more than six times more likely to get COVID. You met this group earlier (Roger & Connie, above).

Persona Analysis Characteristics Used For Segmentation

Striving Forward (NYC):

Striving Forward is a family segment with a mix of single parents and couples with children living in downscale city neighborhoods. Three quarters of residents don’t speak English — many came to these immigrant gateway communities in big cities in the West and Northeast in search of a better life. However, with their below average educations — less than 30 percent have finished high school — these adults tend to earn low wages as blue-collar laborers or service-sector workers. Most can’t afford to buy a home; they tend to live in inexpensive rental apartments in transitional neighborhoods. Almost a quarter of residents change their address every year. Striving Forward have little disposable income, which provides for only modest lifestyles. When they’re not working, these parents look for child-oriented leisure activities. They take their kids to zoos and aquariums, and a big outing is a trip to a theme park. Some admit to not exercising regularly and spend their evenings at home, cooking and listening to Latin music.

Urban Survivors (SF):

Centered in downscale neighborhoods in large and second tier cities, Urban Legacies are older, lower-income households living in aging houses. Most of the householders are over 50 years old and either widowed or divorced. Nearly a quarter did not graduate from high school while a select few have gone on to earn a bachelor’s degree. Many get by on minimum wages from jobs as blue-collar or service sector workers. With household incomes half the national average, these Americans can only afford modest lifestyles in often old housing.

Countrified Pragmatics (Clark County/Las Vegas):

Countrified Pragmatics are mostly couples, over 80 percent being married, living in modern brick homes and double-wide manufactured homes on recently developed lots. Their ages vary from 30s to 60s and about half of the adults have children. Most are high school-educated, blue-collar workers in manufacturing, transportation and construction. Although many are dual income households, incomes are below the national average, making for tight budgets and modest homes. This is the nation’s top segment for mobile home ownership.

More Likely / Less Likely Profile

Another analysis I do is to examine the data most correlated with each area to develop a More Likely and Less Likely list. I find this provides additional context in understanding the people behind the numbers. You can see the items that show meaningful differences between confirmed cases of COVID.


COVID's Second Attack - Employment:

COVID-19 is hitting the communities that disproportionately are poor, where working from home isn't an option. I wondered if the recent spike in unemployment is hitting the same segments that are also indexing higher on confirmed cases. Few states can keep up with the volume of unemployment calls, let alone publish their data in a way to provide insights. Texas did publish one of the more insightful data sets I've seen on recently unemployed.

First, we can see the top five industries in Texas (in April) cutting jobs.

Persona Analysis Characteristics Used For Segmentation

Check out the Texas map to see the top five job types by region. You can also see new unemployment claims by sex -- more women (54%) than men (46%) became unemployed.

Looking at the over 1 million claims (out of about 10 million households) in Texas, it is mostly a different population compared to those more likely to get COVID. As the red circles of the top 10 segments, recent unemployed is reaching up into middle and upper-middle incomes. Examining the indexes, recent unemployment is broadly distributed across all neighborhood segments.

Persona Analysis Characteristics Used For Segmentation Persona Analysis Characteristics Used For Segmentation

Most striking to me is the broad base impact of unemployment. Normally, I see concentrations in zipcode data, such as the 2.00 or higher index of certain segments as you saw in the COVID chart above. But for Texas unemployment data, the highest index was only 1.24, meaning one neighborhood is 24% more likely to have unemployed than you'd expect from the population count.

Two groups index 1.24 . The first is a group called College & Cafe's. Call them Alex and Ann. This is likely where the restaurant and bar job cuts hit hardest. It is a segment of young singles and recent college graduates living in college communities. The other group is a relatively affluent group called Cosmopolitan Achievers. Call them Steven and Christine. They are an Affluent middle-aged and established couples and families enjoying dynamic lifestyles in metro areas. This group losing employment will have a disproportionate impact on the economy.

The US economy is 70% consumer spending. Groups A through H disproportionately drive the economy with their above average incomes and spending. These segments show a lot of green -- that is, above average index on job losses. If Texas if is indicative, what is remarkable is how broad based the recent unemployment appears to be. According to a poll by McKinsey & Company conducted from March 19 to 22 with 1,502 Americans, 40% of households have experienced income reductions. 45% are cutting back on household spending. 75% expect personal finances to be affected for two or more months. I suspect these percents have not improved in the subsequent month. We are just starting to enter "Phase 2" of the four phase economic transition I described in my Guns, Germs & Beer Economy article. Rebuilding the economy won't be easy. Better understanding each segment can help business adjust to changing demand and rebuild faster.


My hope is this data gives you a more personal understanding of the people and communities affected by COVID-19 (both in sickness and in unemployment).

As I noted earlier, part of my broader effort in this analysis is to build a model of the economy in a post COVID-19 world, so we can best understand how to recover and rebuild. Stay tuned for more data and analysis on how our economy and industries will change. I am working with a great group of companies in the Marketing Research industry to share data and develop insights on where the economy is heading and how to accelerate rebuilding.

Recent Analysis

  • 6 May, 2020:
    The COVID Persona Analysis