Skip to content

Household Panel Data Interpreter

Interpret household panel data — penetration, repeat rate, buying frequency — you paste in.

What is the Household Panel Data Interpreter?

The Household Panel Data Interpreter is a free AI skill that reads household panel data — buyer penetration, repeat rate, purchase frequency, and buying-rate figures — pasted from a panel provider like Circana or Numerator, and turns it into a clear read on how a food or beverage brand's buyer base is actually behaving over time. You paste in the panel output and the business question; it returns a plain-language interpretation of what the numbers show about who is buying, how often, and whether they come back, a diagnosis of whether growth or decline is coming from new buyers or existing ones buying more, and the follow-up questions the panel data alone cannot answer. It is built for insights analysts and category teams who receive a panel export and need the story behind it fast, not another quarter spent staring at cross-tabs. Cross-checking the panel read against live food and beverage demand data helps confirm whether a buyer-behavior shift reflects a lasting change or a short-term blip.

Who it's for

  • Insights analysts interpreting a household panel data export
  • Category managers diagnosing penetration or repeat-rate shifts
  • Market research analysts translating panel output for stakeholders
  • Brand teams checking buyer-base health beyond total sales

What you get

  • A plain-language read of your penetration, repeat rate and frequency figures
  • A new-buyer-versus-existing-buyer diagnosis of any change
  • A buyer-base health assessment beyond total sales dollars
  • The follow-up questions the panel data alone cannot answer
  • A flag on any figure that looks unusual or needs a second source
  • A leadership-ready summary of what the panel data means

How to use it

  1. 1Paste the prompt into ChatGPT, Claude or Gemini.
  2. 2Paste your household panel data table and the business question.
  3. 3Answer its clarifying questions about definitions, period and source.
  4. 4Ask it to reframe the read for a specific stakeholder audience.

See it in action

Example input

Pasting 52-week household panel data for fictional 'Brightleaf' cold-brew coffee: buyer penetration 4.2% (down from 4.8% last year), repeat rate 38% (up from 34%), buying rate 3.1 purchases per buyer per year (up from 2.6). Source: Numerator, US. We're confused why total buyers are shrinking while repeat behavior looks healthier — trying to figure out what to tell leadership.

Example output

Household panel read — Brightleaf cold-brew coffee, 52 weeks (Numerator, US)

Headline
This is a shrinking-but-more-loyal buyer base — fewer households are buying Brightleaf at all, but the ones who do are buying more often and coming back more reliably. That combination usually points to a trial or acquisition problem, not a product or satisfaction problem.

What the numbers show
- Penetration fell from 4.8% to 4.2% of households — roughly a 12.5% relative decline in how many households bought Brightleaf even once this year. This is the number that should worry leadership most, because penetration is typically the hardest metric to rebuild once it slips — it means fewer new households are trying the brand.
- Repeat rate climbed from 34% to 38% — of the households who did buy, a larger share came back for a second purchase. This is a genuinely healthy signal: it suggests the product itself is not driving people away once tried.
- Buying rate rose from 2.6 to 3.1 purchases per buyer per year — your existing buyers are buying more often, not just repeating once. Combined with the repeat-rate gain, this tells you retained buyers are becoming more loyal, not less.

The diagnosis
Put together, this is not a story about people trying Brightleaf and being disappointed — repeat and frequency both improved, which is the opposite of what a satisfaction problem looks like. This is much more likely an acquisition or trial problem: either fewer new households are discovering or trying the brand, or a group of former buyers has stopped buying entirely for a reason unrelated to product experience — a distribution change, a price change, or a shift in where they shop.

What this data alone cannot tell you
- Whether the penetration decline is concentrated in lapsed former buyers (people who used to buy Brightleaf and stopped) or in never-tried households (people who were never exposed to it) — these need different fixes: win-back for the former, trial-driving distribution or sampling for the latter. Panel data can usually split this out with a buyer-flow or gain-loss analysis if your provider offers it — worth requesting as a follow-up cut.
- Whether the penetration drop tracks a specific event — a distribution loss at a retailer, a price increase, a promotional pullback — since the 52-week window alone doesn't show you when in the year the decline happened without a sub-period breakout.
- Whether competitors gained the households you lost, or the category itself contracted — this needs a category-level penetration comparison, not just your own brand's numbers.

Follow-up questions to chase
1. Request a sub-period (quarterly) breakout of penetration to find when in the year the decline concentrated — that timing likely points to the cause.
2. Ask your panel provider for a buyer-flow or gain-loss cut to separate lapsed buyers from never-tried households.
3. Pull your own distribution and pricing history for the same period to check for a self-inflicted cause before assuming a competitive or category story.

Leadership-ready summary
Brightleaf's household base shrank about 12.5% this year, but the buyers who remain are more loyal than a year ago — repeat rate and purchase frequency both improved. This points to an acquisition problem, not a satisfaction problem: something is stopping new or lapsed households from trying or coming back to Brightleaf, while the product itself is retaining the buyers it does reach. The next step is isolating whether this is a trial gap or a win-back opportunity, which needs one more data pull before we recommend a fix.

Want me to draft the specific data request to send your panel provider for the buyer-flow cut?

The prompt

Here's the start of the prompt. Download the free bundle for the full, ready-to-paste version — plus the installable Claude Skill and Custom-GPT instructions.

# Role
You are a senior consumer insights analyst who reads household panel data for food & beverage brands. You translate penetration, repeat rate, and buying-rate figures into a plain-language diagnosis, and never treat total sales dollars as a substitute for buyer-base health.

# Context I'll provide
- Panel data, pasted as text: [PANEL DATA — penetration, repeat rate, buying rate, buyer counts]
- Source and period: [SOURCE + PERIOD e.g. Circana, Numerator, 52 weeks]
- Prior-period comparison: [COMPARISON DATA]
- Business question: [BUSINESS QUESTION]
- Audience: [AUDIENCE]

# Your task
1. If the panel data, period, or business question are missing or vague, ask up to 3 clarifying questions BEFORE writing anything.

Frequently asked questions

What is household panel data?
Household panel data tracks a representative sample of the same households over time — recording what they buy, how often, and whether they repeat — from providers like Circana or Numerator. It answers questions scan data cannot, such as buyer penetration (what share of households bought at all) and repeat rate (how many came back). This skill interprets panel figures you paste in and turns them into a plain-language read on buyer-base health.
How is this different from the Syndicated Data Storyteller skill?
The Syndicated Data Storyteller works with store-level scan data — dollar share, velocity, and distribution figures that describe what sold at retail. This skill works with household panel data, which tracks the same buyers over time and answers different questions entirely: how many households buy at all, how often they repeat, and whether growth is coming from new buyers or existing ones. Scan data tells you what sold; panel data tells you who bought it and whether they came back. Use the storyteller for a scan-data narrative, and this skill when your input is panel-level buyer data.
Which AI models can run this prompt?
Any capable chat model — ChatGPT, Claude, or Google Gemini. The prompt is model-agnostic, so paste your panel data directly into a chat, save the prompt as a Custom GPT, or store it as a reusable skill so every panel export gets read with the same new-buyer-versus-existing-buyer discipline.
Will it invent numbers if my panel export is incomplete?
No. It works strictly from the figures you paste in and will not invent penetration percentages, buyer counts, or comparison-period data that isn't provided. If your export is missing a metric needed to fully diagnose a shift — for example, a buyer-flow cut separating lapsed from never-tried households — it will name that gap explicitly and tell you what to request from your panel provider rather than guessing at the answer.

Related skills

Want the live data behind sharper outputs?

These skills get better with real-time F&B intelligence. See what Tastewise can do for your team.