The secret to unlocking growth could be in your customers’ first names

True story.

I recently spoke with a founder who told me that, last year, his team thought females were particularly interested in their brand, only to realize that they were merely responding to customer surveys more often.

It’s a cautionary tale.

Don’t assume the demographic data you collect in a customer or post purchase survey will match your actual customer base.

Chances are it won’t.

If you’re worried a selection bias is compromising your surveys, then do what the Founder did.

Get first names from your sales data (you only need 1,000 or so, not the full list) and use a service like Namsor to find the probability of each name being M or F. (Aside from a few exceptions such as Casey, Riley, and Jessie most names will have a probability close to 100%.)

Determining age is tougher.

But like gender, you can use your customers’ first names to do it.

Head over to the Social Security Administration and grab an actuary table and a list of every name to have ever appeared on a social security card.

With these datasets, you can calculate the median age of all living people with any given name and gender, and then use the output to find the median age of your customers.

Take “Debra” as an example. The median age of living Debras is 64, and there's a 50% chance any given Debra is between 67 and 59 years old. Some names (like Sam) have a wider age distribution. So while it’s significantly harder to pinpoint age than gender, you can still use this analysis to spot differences between your customers and those who complete your surveys.

Here’s an example from some real client work.

By the way, I created a python script to handle all the calculations… and I don’t know python.

Know who does?

ChatGPT.

Here’s the prompt I used to get things going.

Using the datasets in my downloads folder “Names” and “Mortality_table,” write a python script that estimates the number of people for any given name and gender who are still alive (e.g., Alex, M.). Then, find the median age for any given name and gender (e.g., Alex, M, 37).

I’m often asked, “How do I know if I’m leading the witness?” Yet the real concern isn’t just about the questions you ask but who answers them.

Do your respondents reflect your broader customer base?

Want to use that script to check?

It now comes standard on all $45 Survey Roasts. All you have to do is book and Roast and I'll respond within 24 business hours (M-F) to confirm your Roast.