If you're an ecomm brand, you've got sales data. This is a powerful way to leverage it.
The map below shows data from over 350,000 orders. I mapped each order's ZIP code to its corresponding core-based statistical area to show the percentage of total orders relative to the population of each area. It helped the client assess their geographic penetration heading into a brand awareness campaign — and confirmed their intuition that reach was strongest on the coasts.
This analysis replicates the "Duplication of Purchase Law," as popularized by Byron Sharp in How Brands Grow. It shows how mass market brands tend to sell to the same customer base with few skews from the category norm.
Does this pattern hold in your category — or is it more segmented? The data below comes from 149 shoppers answering: "Which non-alcoholic beverage brands have you drank in the last few weeks?"
Social Explorer is excellent for visualizing demographic data. They provide interactive data maps going back to 1790. I've found it most useful for tracking where people move as they age.
The example below shows shifts in age demographics by census tract, specifically college graduates in Atlanta moving from campuses to urban neighborhoods. This type of analysis is particularly useful for clients running out-of-home advertising across metro areas.
I typically open reports and presentations with custom square pie charts that help viewers grasp the scale of their market and identify key segments, before any survey data enters the picture.
Use Census data to anchor the base population, then layer on survey findings. For a client in the smart thermostat industry, I used Census.gov to establish the number of U.S. households (131 million) and those earning at least $50,000 annually (84 million), then overlaid our survey result: the share of those households that already owned a smart thermostat, translating to roughly 39 million.
A few minutes of manual work in Datawrapper to color the squares — and the TAM becomes the first slide of the deck. The visual below is from a self-funded project I did to find how many people bring their phones with them to the bathroom.
If you're interested in how shoppers spend money, a useful move is to categorize expenditures using the 14 classifications from the US Bureau of Labor Statistics. Incorporating these into your surveys lets you show how individual spending aligns with — or diverges from — the average household.
The example below shows that among single, married without kids, and married with young children households, those with young children spend nearly twice as much annually as single households. Not a surprising finding, but a powerful one to illustrate with BLS benchmarks behind it.
Many brands track significant life transitions as shoppers move from their 20s to their 30s — disposable income increases, homeownership rates rise. But one of the most consequential shifts is starting a family.
The chart below shows the percentage of women aged 18–30 who have had at least one child, overlaid with data from a fitness brand survey showing a positive correlation between fitness app usage and having kids. I've found this approach works across a surprising range of categories.
Many ecomm brands know their customers well but lack clean age data. Tools like Google Analytics give partial signals — but sales data almost always includes first names, which you can use to estimate age and gender.
This is especially useful for checking whether respondents in a customer or post-purchase survey reflect your actual customer base. The dot plot below shows 17 names randomly selected from a database of over 100,000 first names on Social Security cards dating back to 1881.
One way to segment your market is to identify the central rub in your category — the behavior people think about, worry about, or quietly obsess over.
Then ask two simple questions: one about frequency ("How often do you…") and one about emotion ("What's your typical reaction when you…?"). Plot the answers in a heatmap.
The example below — created for a personal finance brand — looks at how often people check their bank account and how they feel when they do. What we found surprised us. A meaningful segment felt far more content about their finances than we'd assumed — a group we hadn't seriously considered targeting.
This is one of the most presentation-ready outputs you can get from a survey.
The idea: pick two dimensions that define the tension in your category (premium vs. accessible, functional vs. aspirational, niche vs. mass market), ask respondents to rate each brand on both axes, then plot the results on a scatter chart. The chart becomes the presentation slide. No additional design work needed.
The visual below shows brand currency scores for smart home fitness brands mapped across perceived momentum and brand identity. Brands cluster in ways that are immediately legible — and immediately useful for positioning conversations.
This framework is inspired by Richard Thaler's distinction between acquisition value and transaction value. The core question: does your product feel worth it — not just based on price, but on how long people expect it to last?
Ask respondents to estimate how long a product will last and what they'd expect to pay for it. Plot both answers together and you get a read on perceived value that goes beyond price sensitivity. Products in the high-lifespan, low-price quadrant are ripe for premium repositioning. Products in the low-lifespan, high-price quadrant have a trust problem.
This chart works especially well for consumer goods, durables, and any category where "worth it" is a common purchase barrier.
Most survey questions about preferences produce boring, predictable results. "How satisfied are you with…" on a 1–5 scale doesn't tell you much.
This one does. Ask respondents to rate a list of items — brands, products, moments, experiences — as overrated, about right, or underrated. The framing forces people to take a position relative to the mainstream view, which produces opinions that are sharper and more honest than standard satisfaction ratings. It also makes for a more interesting chart.
The example below applies this to Super Bowl items by fan base — but the format works for any category where you want to know what people actually think, not just how much they like something.
Open-ended survey responses are easy to collect and hard to use. This approach turns them into a visual that actually tells a story.
Ask two groups the same open-ended question, extract the most common words or phrases from each group's responses, then visualize the overlap and divergence between them. The bubble chart below shows words used by social media users who said the platform is "mostly bad" for society versus those who said it's "mostly good" — even though both groups used social media that same day. The words people reach for to explain the same behavior reveal the gap between their stated attitudes and their actual habits.
This technique works any time you have two distinct groups responding to the same prompt: buyers vs. non-buyers, loyalists vs. churners, satisfied vs. dissatisfied.
