Improving Data and Creating Pandemic Personas

July 29, 2020

Analytics IQ examined some of the causes and consequences of bad data and then moved on to identify four personas that elucidate the different ways in which Americans are responding to the COVID-19 pandemic.

Analytics IQ pointed out three reasons why marketers can find themselves working with bad data.


  • They rely on customers to update their own data. Customers, however, can provide inaccurate data — for instance, when they give themselves a younger birthday or when they supply an old email address.
  • They rely on digital behavior data, which can also be misleading. The person shopping at Anthropologie online might be a man shopping for a Christmas gift for his wife and not a true loyalist of the brand.
  • They rely on disparate data sets. This can create the potential for confusion, not to say inconvenience, when these data sets need to be merged, especially if different business units have different definitions for the same data.

Analytics IQ elaborated on the problem of bad data by laying out some of the factors that can make it bad in the first place.

  • Poor identifiers: A wrong name can cause data to be mismatched.
  • Low volume: Having a limited perspective due to limited data can cause marketers to draw unjustified inferences. Someone shopping at Baby's 'R' Us might merely be a guest at a baby shower and not herself expecting.
  • Data latency: It can be difficult to collect and analyze data fast enough to stay ahead of a consumer before he or she is out of the market. Once that exit is made, the data may not be of any use, though the marketer may not know this.
  • Untested data: Unvalidated data is probably unreliable.

If marketers are working with bad data, it can hurt their bottom line in a number of ways.

  • Media waste: Bad data results in media served to consumers who are not the brand's actual target.
  • Opportunity cost: The marketing resources mistakenly devoted to off-target consumers could have been used on on-target ones.
  • Over-priced data: If the data was paid for, the percentage of inaccurate data represents wasted money.

Analytics IQ urged marketers to avoid these data pitfalls by basing their analytics on offline data (e.g., name, address, etc.). As an example of the kind of results that can be realized with this kind of data, the company showcased its pandemic personas, which it had created using its own proprietary research.

These personas describe four sets of responses that consumers are mounting to the COVID-19 pandemic. They take into account both the consumers' attitudes and the protective actions that they are taking.

The Careful Optimist is taking precautionary measure but is maintaining a positive attitude. He or she is prioritizing purchases of masks, gloves, and herbal self-care products. Members of this persona group are coping through activities such as journaling, meditating, and calling family members. They are also:

  • The oldest of the four persona groups
  • Two-thirds female
  • Two-thirds single
  • In possession of only a high school education in over half of cases
  • More likely to be Democrats
  • Likely to live in suburbs or cities
  • Distinguished by below average income and wealth

The On-Guard Cynic also takes precautionary measures against the virus but has a negative attitude. People who belong to this persona group are prioritizing purchases of soap/sanitizer, masks, and over-the-counter medications. They are coping by consuming alcohol, cleaning, and playing music. In addition, they are:

  • Twice as likely to belong to Gen Z
  • Two-thirds female
  • 70 percent single
  • Marked by the highest proportion of college graduates of all of the persona groups
  • Marked by the highest concentration of Democrats
  • Most likely to live in big cities
  • Marked by the highest level of income and wealth of all four personas

The Lax Dreamer eschews protective measures and maintains a positive attitude. Members of this persona group are prioritizing purchases that include household sanitizer, groceries, and over-the-counter medication. They are coping by praying, playing music, and listening to music. In addition, they are:

  • In 82 percent of cases, between the ages of 45 and 64
  • Two-thirds male
  • Marked by the highest marriage rate of all four persona groups
  • In possession of a high school education in the majority of cases
  • Almost twice as likely to be Republican
  • Twice as likely to live in small towns or on farms
  • Distinguished by the lowest level of income and wealth of all four persona groups

The Remiss Pessimist combines a failure to take precautionary measures with a negative attitude. Members of this persona group are prioritizing purchases that include alcohol, DIY projects, and apparel. They are coping by consuming alcohol, eating, and cleaning. In addition, Analytics IQ found that they:

  • Skew younger than other groups
  • Are men in over 82 percent of cases
  • Are distinguished by a below-average divorce rate
  • Are highly educated
  • Are represented by more political independents than the other groups
  • Are easily found throughout all communities
  • Are the most impulsive spenders

Q&A with Zach Hudson, VP of Solutions at Analytics IQ

Q. What are some data points that marketers should be extra cautious of?
A. One thing that will give us some trouble is if a client has an email address as the main identifier in their data. So many people have multiple email addresses. Sometimes they don't remember which ones they entered for which site. And sometimes they give you an old one. That can be a barrier to correctly identifying a record.

Q. Does Analytics IQ have B2B data?
A. Yes. In addition to our consumer database, we have a B2B database that contains roughly 24 million businesses. And then we also have 65 million business contacts on that file.

Q. How can you convince management that data validation is important, though it can be costly and labor-intensive?
A. For most companies, the proof is in the pudding. The failure of projects that didn't validate their data will convince them of the importance of validating going forward to get the correct, actionable insights they need.

Q. Can your data be used digitally?
A. Our data lives offline. But through partnerships with Neustar and LiveRamp, we have the ability to send data online so it can be activated across channels for digital, social — pretty much anything.

Q. Is the granularity of your person data on the individual level? If so, how do you ensure privacy for individuals?
A. As of now, a large number of our data points are at the individual level. But because of the way our data is pulled together, its compliant. All of our data is inferred, so very little of our data is strictly known. It's usually a combination of known and modelled data.

We're also looking at the possibility of rolling up some of our data points to the household level as a response to the market and to protect ourselves against future privacy regulations that might come down the road.

Source

"Improving Data and Creating Pandemic Personas." Zach Hudson, VP of Solutions at Analytics IQ. ANA 1-Day Conference: The Future of Consumer-Enhancing Data and Analytics, 7/29/20.

You must be logged in to submit a comment.