Glossary

Panel data

Panel data, in consumer goods, is purchase information reported over time by a standing group of households that agree to share what they buy and may be rewarded for taking part. Research firms weight the results, adjusting them so the group matches a larger population such as all households in a country, and sell them to brands as estimates.

How it works

A research firm recruits a standing group of households and asks them to report what they buy, week after week, for months or years. Reporting methods vary by provider and include barcode scanning, receipt uploads, linked digital accounts, purchase diaries, and surveys. The firm then weights the results, which means it adjusts the counts so the group matches a larger population by traits such as household size, income, and region. That is how a sample of households can be used to estimate a country's buying, within a margin of error.

A brand buys reports built from that group. It receives percentages and projected totals, such as how many households bought its product in a quarter. Because the same households are followed over time, a panel can also show whether households buying for the first time in the panel's records bought again, which a count of register sales cannot show.

In statistics, the same two words mean any dataset that measures the same people, firms, or places repeatedly over time. Consumer panel data applies that idea to household shopping. Panel data is one kind of syndicated data. A second kind is built from store sales feeds, covered under POS data.

Why it matters

Panel data tells a brand how many households buy its product and how often, across the stores those households report. A register-level count shows what sold in each store, and on its own does not say which households bought. A panel fills in that household view. Numerator's panel guide lists household penetration and buying rate among the measures panels report, and Circana lists penetration, repeat buying, and brand switching, meaning shoppers moving from one brand to another. Repeat buying is one way to read retrial, a household's next purchase after its first one, which is called trial.

A panel is a sample, so its figures are estimates, and it covers only the households that take part. NielsenIQ lists representativeness as a limit, because a recruited panel can miss parts of the market. A brand can see the trend in its reports but cannot see which households are in them. For one way a brand can learn who its own buyers are, see our post on how a receipt scanning loyalty program works.

Example

A panel report can show a brand that new buyers are not coming back. A sparkling water brand sees household penetration rise by two percentage points over a year, while the share of its buyers who purchased again within that year falls. Together the two readings may point to more households trying the product without making it a habit, though the mix of buyers also changes when many new households try something once. The report shows the pattern but cannot say which households they are or why they stopped, so the brand would have to ask its own customers.

Common mistake

A brand can read panel figures as exact counts when they are estimates from a sample. For a small brand, a small region, or a short period, a change of a few tenths of a point can come from a handful of households. Ask the provider for the sample size behind the number and compare several periods before acting on a single reading.

Where TrueLoyal® fits: With receipt scanning, which TrueLoyal turns on for an account on request, members can submit receipts for offline purchases, and the store, products, and purchase date are read from each receipt and saved with the transaction. That view covers only members who scan, so it sits alongside panel data and the brand's other syndicated data.

Frequently asked questions

What is the difference between panel data and POS data?

POS data counts what sold at the register, store by store, and covers the stores whose sales data the brand can get. Panel data follows what a sample of households bought across stores over time, so it shows household behavior such as who bought and whether they came back.

How is consumer panel data collected?

Methods vary by provider and include barcode scanning, receipt uploads, linked digital accounts, purchase diaries, and short surveys. The provider then weights the results so the group of households matches a wider population. Circana, one panel provider, says it also calibrates its panel against retailer sales data.

Can a brand contact the households in a panel?

A brand cannot contact them directly. Research firms protect panel members' privacy, so a brand receives reports built from the panel and has no list of households to email or reward.

Is panel data the same as loyalty program data?

No, because panel households are recruited to stand for a wider population, while loyalty members are people who chose to join one brand's program. Panel data describes a market, and loyalty data describes one brand's members.

Related terms

Retail and category vocabulary
Syndicated data

Syndicated data is market data that a research company collects once and sells to many subscribers. In consumer goods it usually means retail sales data and household panel data from providers such as Nielsen and Circana. It is the shared source of truth that brands, retailers, and brokers use when they talk about category performance.

Retail and category vocabulary
POS data

POS data is the record of sales captured at the point of sale, which is the register or checkout where a transaction happens. It shows what sold, in what quantity, at what price, in which store, and when. For brands selling through retail, it is the most direct count of how many units actually moved to shoppers.

Retail and category vocabulary
Household penetration

Household penetration is the percentage of households in a market that bought a brand at least once in a given period. It answers how many people are buying rather than how much they are buying.

Retail and category vocabulary
Buying rate

Buying rate is how much each buying household purchases of a brand or category over a set period, usually a year. It is measured in units, dollars, or purchase occasions per buying household. Together with household penetration it explains where a brand's volume comes from, because penetration times buying rate equals volume.

Retail and category vocabulary
Trial

Trial is a household's first purchase of a brand or product. It is the moment a shopper moves from never having bought to having bought once, and it is what grows household penetration. In TrueLoyal®'s vocabulary trial is the first buy and retrial is every buy after it.

Retail and category vocabulary
Retrial

Retrial is every purchase a household makes of a brand after its first one. Many teams call it repeat purchase, and it is usually discussed together with trial as trial and retrial. Retrial drives buying rate, which is how much each buying household purchases over a period.

Close
Schedule a Demo