Retail market research starts with Census data and ends in the aisle

Retail market research answers one of two questions: will this product sell in this kind of store, or will this location support this store? Free Census data shows how the channel is trending, store visits show what is on the shelf today, and conversations with retail buyers, store managers and merchandisers explain how products get onto that shelf and why they leave it.

The data and the visits are cheap. The conversations are where most of the useful answers are, because shelf decisions are made by people, and those people rarely publish their reasoning.

1. Write the decision as a retail question first

A category question names the product, the chain type and the region. A location question names the store format and the specific sites. "Is the snack market growing?" is neither.

Here is a made-up example used throughout. Dana runs a small hot sauce brand. It sells at farmers markets and in 12 independent grocery stores, and a regional supermarket chain has invited her to pitch. Her decision: pitch now with three flavors, pitch now with one flavor, or stay in independents for another year.

Every step below should feed that one sentence.

2. Census retail data shows the channel trend, not your product's chances

The Census Bureau runs several free retail programs. They are good for checking that a channel is not shrinking under you, and for sizing a category in a state. None of them will tell you whether a buyer wants your product.

The Census also publishes an experimental Monthly State Retail Sales release with modeled state-level figures. Use it as a rough regional check, and say it is modeled when you quote it.

The Economic Census product lines are the most useful for a category question. The Census Bureau's own example shows California grocery stores selling nearly $5.2 billion of frozen foods in 2012, and meat and poultry making up 16.1% of Louisiana grocery store sales. For Dana, the same tables can show how much of a state's grocery sales fall in her broad product line. Check the product codes before you quote anything: a broad line such as condiments is not her market.

Short version: use Census data to rule out a shrinking channel or a tiny region. Do not use it to forecast your own sales.

3. For a location, count competitors first and treat foot traffic as an estimate

If your question is a site, start with counts. County Business Patterns gives establishments by industry and county, and ZIP Code Business Patterns gives establishment counts at the ZIP code level, including employment size classes for detailed industry codes. From reference year 2019 the ZIP code data sits inside the County Business Patterns data tools.

Then look at trade area and foot traffic data, which comes from commercial vendors. The main types are demographic estimates for a drawn area around a site, traffic counts on nearby roads and visit estimates from phone location panels. That last kind works like a survey. Placer.ai, one of the larger vendors, says its data comes from tens of millions of devices through hundreds of partner apps, a panel it describes as about 7% of the US population, and that it runs a proprietary extrapolation algorithm to estimate total visits (Placer.ai). So every visit count is a projection from a sample. Compare sites with the same vendor, and do not mix numbers across vendors.

Counts and visits still miss whether people near the site buy what you sell. The walking, watching and catchment drawing in our restaurant market research guide apply to shops too.

4. Visit the stores and record what the buyer already decided

The shelf is a public record of past decisions. Visit at least three stores of the target chain, at different times of day, and write down the same things in each:

• Facings. How many units of each product sit at the front edge of the shelf. Read more facings as a bigger bet by the retailer.
• Height. Which brands get eye level and which sit at the bottom.
• Price points. The cheapest, the most common and the most expensive item in your segment, per ounce.
• Private label. Whether the store brand sits next to the category leader, and how far below it is priced.
• Gaps and tags. Empty slots, clearance stickers and discontinued tags. These are products on their way out.

For Dana, the visit answers practical questions. If the chain carries eight hot sauces in a single four-foot section and two are on clearance, a new brand is competing for one of two slots that are about to open.

What a visit cannot show is why. A product at the bottom shelf might be failing, or might be new, or might be the brand that declined to pay for better placement.

5. Ask buyers, store managers and merchandisers how the shelf is decided

Three roles see different parts of the decision:

• Buyers and category managers at headquarters decide what the chain carries, what it costs to get in and when an item gets cut.
• Store managers see whether products actually reach the shelf, how often they run out and what shoppers ask for.
• Merchandisers (often working for a brand, a broker or a distributor) reset shelves to the planogram, the diagram that says where each product goes, and know how closely stores follow it.

Talk to former buyers before current ones. A current buyer at your target chain hears your call as an early pitch, and their answers will be polite. A former buyer in the same category can tell you how reviews worked without anything at stake.

Ask about slotting allowances. The FTC defines them as one-time payments a supplier makes to a retailer as a condition for initial shelf placement or warehouse access. Its 2003 staff study of five grocery categories found considerable variability across categories in whether fees were paid and how much. It also found they were less frequent and lower for products delivered directly to stores rather than through the retailer's warehouse. The study used a small set of anonymous case studies, so treat it as background and ask someone who has negotiated these recently.

Questions worth asking a former buyer:

1. How was a new item in this category judged in its first year, and by what measure?
2. What did it cost a small brand to get in, beyond slotting: free product for the first shelf fill, promotions, in-store demos?
3. What usually got an item delisted, and how much warning did the brand get?
4. Did store-delivered brands get treated differently from warehouse-delivered ones?
5. What would make you choose a three-flavor launch over a single flavor?

Ask store managers the shorter version: what sells without a promotion, what keeps running out, and what gets pulled first when the planogram resets. Do not ask anyone for their employer's confidential sales figures. How they made decisions is fair game; a chain's internal numbers are not, and the channel checks guide covers where that line sits.

6. Decide what result means go before you collect it

Write the threshold down before the calls. For Dana, it could be: pitch now if two former buyers say a three-flavor launch is normal for a new brand and the cost of entry fits what she earns from her independents in a year. Pitch with one flavor if the buyers say a single flavor is the usual test. Stay put if entry costs exceed that year of profit.

Without a threshold, every answer sounds encouraging. If you are working through this on a small budget, market research for small business covers choosing the cheapest evidence that could change the decision.

Your next step

Pull one Census table for your channel, visit three stores this week and write down the facings and price points. Then get two former buyers from your category on the phone. Instant Expert can find people matching a description such as "former grocery buyers who managed the condiments category at a regional supermarket chain," send each a paid invitation for a short call, and you pay only for calls that get booked. Start with the directory pages for retail buying experts, shelf placement and slotting fee experts and retail site selection experts.