At 3Q Digital, we’re always pushing the innovation envelope; we believe this is an essential component of success in an industry that changes so frequently. This post is part of our inaugural Innovation Week, where we showcase all manner of innovations that have improved results for our clients and teammates.

These days, information on your customers, their demographics, interests, and behaviors, is extremely accessible. However, in-depth customer insights almost always come with a fee. In most cases, the cost of knowing more about who buys is worth the investment. But when budget is tight, do not fear. Simply consult the government for visibility into characteristics defining your customers

By government, I mean the US Census Fact Finder. The United States Census Bureau kindly aggregates Population, Demographic, Housing, Economic, Social, and Education characteristic estimates for the likes of public use.

This characteristic data can be cut by geography, with state, county, and zip code being just a few of the data modification options.

So how does this apply to your customers? Well if your customers are U.S. residents, and you collect zip code data with transactions, then you’re in luck. In minimal time, thanks to the Census data and your preferred stats application, you can obtain a breadth of information on your customers.

Here’s how:

Step 1. Navigate to the US Census Fact Finder and click the Annual Estimates of the Resident Population link.

Below is a glimpse of what that page looks like. As you can see, modifying the table and adding geographic specifications are featured actions (making it that much easier to do).

Step 2. Add the Geographic specification and download the data.

I recommend the 5-digit zip code. Once you add the 5-digit zip code to your selection, select “Show Table.” Then select download.

Step 3. Download a customer transaction file.

In order to learn more about the buyers, you will need to cut the transactions by Zip Code.

Step 4. Head back to US Census Fact Finder and download additional characteristic files.

There are a number available; housing, social, and economic characteristics are a good start. Cut those files by zip code, as you did for the population estimates, and download.

Step 5. Use your preferred stats application to run correlations on characteristics.

Ingest the census data and sales data by zip code into the stats application and join the two data sets by zip code. Then create a normalized variable of success called “Transactions per Capita” using the population data provided by the census and transaction data. Once this variable is established, loop through each census characteristic and calculate the correlation between the census characteristic and the normalized measure of success to understand if there is a positive or negative correlation (or no relation) between these variables.

Step 6. Isolate leading characteristics correlating with transactions per capita.

Do your consumers correlate more with the renters characteristic than the home owner characteristic? Are they more likely to work from home than commute? Are they most likely to fall in the Household Income bracket of $75k-99k? The correlation coefficient will reveal a more detailed profile on who your consumer is. However, proceed with caution when using correlation coefficients in analysis. There are cases when correlations exist but causation does not. Tyler Vigen gets to the root of that issue, with the example of Divorce Rate in Maine correlating with Per Capita Consumption of Margarine.

Step 7. Adjust media plan based on findings.

There are a number of ways to do this. Here are a few examples:

  • Zips with Organic Interest: Identify these markets by locating zips with highest Network per Capita. As long as these zips were not served marketing that other zips were not, organic interest is revealed.
  • Incorporating Leading Characteristics into Targeting Strategy: Layer in leading characteristics among purchasers into targeting strategy. For example, if your customers are renters who work from home, target zips that index highest on those characteristics or specifically target those characteristics.

You can get creative here. But do keep in mind Vigen’s warning.

Enjoy the free insights!

Leave a Comment

Christina Oswald
Christina joined the 3Q NYC Analytics team in August 2016. Prior to starting at 3Q, Christina specialized in Social Analytics (agency and client side) and Quantitative Market Research. Christina was born and raised in Blue Point, NY (think oysters and beer), and now lives in Brooklyn. She has an abnormal obsession with reading and dogs (especially her puggle, Ginny Judy).