Category intelligence combines three views of a category: what the data shows, what retailers observe and what shoppers say and do. Retailer insights fill the gap that sales data leaves: unmet requests, substitutions and hesitation that no dataset records. Read against the numbers, they turn a change in share into a hypothesis you can size.
Why talk to retailers for category insights?
Retailers observe demand that data cannot see: unmet requests, the substitutes chosen and the hesitation before a lost sale. They also surface early signals, such as a new need or a competitor's scheme, months before a category report shows them. In India, much of this happens across a counter.
NielsenIQ puts general trade at about 82% of FMCG value (2026).[3] Redseer's 2026 report puts kirana stores at about 91% of grocery in 2025.[4] The two measure different things, FMCG value against grocery spend, but both point to the counter.
What is category intelligence?
Category intelligence is the combined view of a category from three perspectives: what the data shows, what retailers observe and what shoppers say and do. It goes beyond category insights because every retailer theme is checked against the numbers before anyone acts on it.
What are retailer insights?
Retailer insights are what retailers, buyers, store staff and distributors observe about shoppers and the category. They cover the brands requested by name, the substitutes taken when a brand is missing, and the doubts that end a sale. Most of this never reaches a sales report.
Retail measurement and panels show what sold, where and at what price, the foundation of most CPG market research. One retailer's view is a lead, not a verdict.
What is category management?
Category management is a retailer and supplier process that runs a product category as a strategic business unit. GS1 and ECR describe it as giving the category a defined role and tactics such as assortment, placement and promotion.[5] Category management insights persuade a buyer most when they show what shoppers asked for, substituted and could not find.
Retailer interviews work best as a designed programme
A retailer insight programme speaks to several kinds of trade respondent, across formats and town classes, at a regular rhythm. Spread across them is what protects against one-store bias.
What are channel checks and trade surveys?
Channel checks, trade surveys and retailer surveys are names for the same idea: structured conversations with retailers, distributors and salesmen about what sells, what is missing and why. A retailer insight programme adds design: who to speak to, where and with which questions.
Speak to kirana owners, modern-trade store managers and buyers, quick-commerce category managers, distributors and salesmen. Add former category or sales managers, who explain what similar patterns meant before.
| Respondent | Sees first hand | Bias to correct for |
|---|---|---|
| Kirana owner or counter staff | Brand requests in shoppers' own words, substitution at the counter, credit and pack size needs | Margin and scheme preferences; the owner's own brand habits |
| Modern-trade store manager | Shelf gaps, complaints, promotion response, footfall by day part | Store-level view; limited sight of central decisions |
| Modern-trade buyer or category manager | Range review logic, category role, supplier performance, trade terms | Commercial interest in terms; speaks for the chain, not the shopper |
| Quick-commerce category manager | Search terms, null searches (searches that return no product), basket pairs, dark-store stocking decisions | Platform view; metro skew; commercial interest in ad spend |
| Distributor and salesman | Which retailers order what, credit, scheme response, competitor push | Own margins and targets; may describe retailers rather than shoppers |
| Former category or sales manager | How the category has behaved over cycles, what moved share before | Memory of an earlier market; check against current data |
How many retailer interviews, where and how often?
As Nextyn's working guide, 30 to 60 retailer interviews across formats and regions are often enough for one category in one country. The test is whether new interviews still produce new themes. A 2022 systematic review found that fairly uniform groups reached saturation within 9 to 17 interviews.[6]
Spread matters more than the total. A category can behave very differently across metro, Tier 2 and Tier 3 towns, and across regions. A grid that crosses format with town class and region keeps one environment from dominating. Weight the grid toward the formats and regions that carry most of the brand's sales.
On timing, a one-off study suits a single decision such as a launch. A quarterly or half-yearly wave with the same core questions shows how requests and substitution change.
Three retailer interview methods: brand requests, substitution and objections
Three retailer questions become three methods: brand request logging, substitution path mapping and objection capture.
What questions should you ask retailers about a category?
Ask what shoppers do at the counter or the shelf, in three questions. Which brands do they request by name, and in what words? What do they take when that brand is missing? What makes them hesitate before choosing, or leave without buying?
What is brand request logging?
Brand request logging asks the retailer to recall, in rank order, the brands shoppers ask for and the words they use. Brand requests measure pull. A brand that sells well but is rarely asked for by name is being chosen at the counter. That leaves it exposed to substitution.
What is substitution path mapping?
Substitution path mapping records, for each brand and pack, what shoppers take when it is missing, by step of distance from the original choice. It reveals the real competitive set. A shopper who moves to another brand in the same pack size tells you something different from one who changes format or leaves.
| When the preferred item is missing, the shopper | What it tells you | Typical response for the brand |
|---|---|---|
| Takes the same brand, another pack or variant | Brand loyalty is strong; pack architecture provides cover | Protect the core pack |
| Takes another brand, same format and price point | The competitor at that price is the real rival | Availability and visibility first; then a clearer reason to ask by name |
| Takes a different format for the same need | The category boundary is wider than the brand assumed | Redefine the competitive set; review the range across formats |
| Goes to another store or channel | The brand pulls; the retailer loses the trip | Use as evidence in retailer conversations |
| Delays the purchase | Loyal and patient; low immediate loss if restocked quickly | Restock speed matters more than shelf presence |
| Does not buy | Need is weak or the category is discretionary | Investigate relevance and value, not only availability |
Retailers can usually name the specific substitute, which makes the matrix a practical form of retail competitor analysis.
What is objection capture?
Objection capture asks retailers which doubts shoppers raise, about which brands, and what happens next. Kantar finds about 30 to 35% of decisions remain open at the shelf.[1] Retailers notice these doubts long before a dip appears in a category report.
Out-of-stock substitution shows the real cost of a missing brand
What a gap costs a brand depends on where its own shoppers go, which retailer interviews can reveal.
What do shoppers do when their brand is out of stock?
Most shoppers switch rather than wait: in a 29-country study, 31% went to another store, 26% bought another brand and only 15% delayed.[7] The study, by Gruen, Corsten and Bharadwaj for the Grocery Manufacturers of America, FMI and CIES, is the most widely cited on the subject.
It covered about 71,000 shoppers and found an average out-of-stock rate of 8.3%.[7] Corsten later noted that 72% of root causes sat in the store itself, in ordering, forecasting and shelf replenishment.[8]
NielsenIQ reported in 2022 that 70% of shoppers switch brand when their usual one is out of stock. The article gives no base, sample or method for that figure.[2]
- Buy at another store31%
- Buy a different brand26%
- Same brand, other size or variety19%
- Delay the purchase15%
- Do not buy9%
Source: Gruen, Corsten and Bharadwaj (2002), summarised in Corsten and Gruen, Harvard Business Review (2004).[7]
How much do out-of-stocks cost retailers and brands?
In 2022 about 4% of FMCG products were out of stock on shelf in France, Spain and the UK. NielsenIQ linked rates of 4.4% in France and 4.0% in the UK to €851 million and £1 billion in missed sales.[2] IHL estimates that inventory distortion, mostly out of stocks, cost retailers about $1.73 trillion in 2025.[10]
McKinsey's US Consumer Pulse survey of October 2021 (n = 2,095) asked about out of stocks. Of those who could not buy what they wanted, 39% switched brands, 32% went to another retailer and only 13% waited.[9]
None of these studies covers Indian general trade, so the shares are not Indian figures. The direction matches Kantar's 2026 view for India: an out of stock reopens a decision that was otherwise closed.[1]
In general trade, availability depends on how the distributor and salesman serve the outlet: visit frequency, credit and the ease of reordering a single pack.
How to prioritise availability: the out-of-stock exposure grid
| How often missing | Shoppers mostly delay, take another pack or switch store | Shoppers mostly take another brand |
|---|---|---|
| Often out of stock | Fix supply, protect pull. The retailer loses trips. | First priority. Every gap hands volume to a named competitor. |
| Rarely out of stock | Maintain. The brand is protected by pull and presence. | Build reasons to ask by name. The brand has little cover if availability slips. |
- Fill the substitution matrix for each priority SKU before setting availability targets.
- Place each SKU on the exposure grid, using retail audit data or retailer recall for how often it is missing. Fund the top-right cell first.
- Ask retailers and salesmen why the SKU goes missing: ordering, credit, visit frequency or supply. If it is on the shelf but still not selling, use the slow seller diagnostic for retail audits instead.
- On quick commerce, check availability in each dark store, not only the platform listing.
Kirana shopkeeper influence shapes brand choice most in new categories
Kirana owners influence brand choice, but less than often assumed.
Do kirana owners influence brand choice in India?
Yes, but selectively. Kantar Shopper Intelligence (September 2026) reports that the retailer retrieves the product and places it on the counter in about 90% of purchases. Retailer advocacy sways about 10 to 15% of purchases in established categories and 30 to 40% in new or unfamiliar ones.[1]
| Shopper state on entering the store | Share of shoppers | Where influence works |
|---|---|---|
| Brand already decided | 65 to 70% | About 98% leave with that brand; the main risk is the brand being out of stock |
| Category planned, brand open | About 20% | Shelf findability, reassurance and retailer advice |
| Category reminded in store | Around 15% | Mostly triggered by visibility |
What makes a shopkeeper willing to recommend a brand?
Four things make a shopkeeper willing to recommend: confidence in the product, low risk of complaints, commercial sense and reliable supply, since no retailer recommends what they cannot restock. Retailer interviews show which is missing, so a launch can design its trade offer around it.
For established brands, Kantar's guidance is to treat availability as the first line of brand defence.[1]
How do D2C brands entering general trade or quick commerce use retailer insights?
A D2C brand entering general trade or quick commerce meets a trade it has not sold through before. Advocacy interviews show what a shopkeeper needs before recommending a brand in a new category, where Kantar finds advocacy sways 30 to 40% of purchases.[1] Brand request logging shows whether shoppers ask for it by name or by price point.
Distributor and salesman interviews show whether credit terms and visit frequency will keep a small pack in stock. On quick commerce, category managers can show where a listed SKU is missing from the dark stores that serve its strongest neighbourhoods.
How to combine retailer interviews with panel and retail measurement data
Neither source is enough on its own.
How do you combine retailer interviews with panel data?
Combine them in a loop: let a surprise in the data frame the retailer questions, then size each repeated theme back in the data before acting. Retail measurement and panels show the size and direction of a category; retailer interviews and shopper research explain what drives it.
What can't retail measurement show in India?
Retail measurement and panels are essential for share, distribution, price and trend. In India they have four limits: coverage of general trade and quick commerce, reporting lag, no record of lost demand and no reasons. These are the boundaries of transaction data, not faults of any provider.
- Coverage. General trade is vast and fragmented, and quick commerce is measured differently from stores, a gap that quick-commerce research has to work around.
- Lag. It reports by period, and retailers often notice a shift weeks or months earlier.
- No lost demand. It records what sold, not the shopper who asked for a missing brand and left.
- No reasons. A share decline looks the same whether it comes from price, availability, a competitor's scheme or a change in what shoppers need.
Worldpanel put quick commerce at nearly a third of online FMCG in the year to May 2025. NielsenIQ's Q2 2026 snapshot puts it at about 85% of ecommerce FMCG across India's 52 largest metros. The two differ in geography, period and method: a household panel against retailer sales data.[11][3]
Bain and Flipkart's 2026 report put online grocery at only about 1.5% of grocery.[12] Almost all grocery is still bought offline, much of it in the general trade that is hardest to measure.
The signal, story, size workflow
The signal, story, size workflow starts with an anomaly in the data and asks retailers and shoppers to explain it. It then goes back to the data to size each explanation before acting.
- Signal: read the data first. List what you cannot explain, by region, channel and pack.
- Story: ask the trade. Turn each anomaly into retailer and expert questions.
- Story: ask the shopper. Test shopper reasons with path to purchase research: shop alongs, exit interviews or a short survey.
- Size: back to the numbers. If retailers name a substitute, its share should rise where you are absent.
- Decide and track. Act where the perspectives agree. Run a small test where they do not.
Source: Nextyn analysis.
An illustrative example: retailer interviews explain a share dip
In this invented scenario, retailer interviews explain a snack brand's share dip in one Indian state, and the data confirms it. It describes no real company.
The retailer insight programme covers about thirty kirana stores across four towns, a handful of modern-trade store managers, two distributors, several salesmen and two former category managers.
Share has fallen by a little over a point, concentrated in Tier 2 and Tier 3 towns. Retailers say shoppers still ask for "the 10-rupee packet". When it is missing, most take a competitor's pack at the same price. The small pack has been hard to get since a distributor changed its credit terms. Former category managers add that the competitor has been offering a retailer scheme on its 10-rupee pack in the region.
Back in the data, weighted distribution of the 10-rupee pack has fallen in the affected distributor's territories, where the competitor's share has risen most. The brand restores supply through the distributor, increases salesman visits to small outlets, and checks the competitor's scheme terms against its own. Shoppers move to the competitor, so the small pack sits in the top-right cell of the exposure grid: first priority. Without the interviews, the team might have read the dip as a preference shift and cut price.
Category intelligence terms used in this paper
| Term | Meaning |
|---|---|
| Substitution matrix | A record of where shoppers go when their preferred brand or pack is unavailable: another pack, another brand, another format, another store, delay or no purchase. |
| Weighted distribution | The share of category sales made by the stores that stock an item, rather than the simple count of stores. |
| General trade and modern trade | In India, general trade means independent stores such as kiranas served through distributors; modern trade means organised chains such as supermarkets and hypermarkets. |
How this paper was built
This paper draws on public sources published between 2002 and November 2026, quoted as published. No published study of out-of-stock response in Indian general trade that states its method was found. Kantar's 2026 kirana findings are quoted as reported, without stated sample or method.
The full report (PDF) carries the sampling grid, the question banks and a pharmacy and specialist row for the map above.
Frequently asked questions about retailer insights
Sources: the 12 public sources behind the full report
- 01
Kantar (2026). Maitra, S. The myth of the kirana shopkeeper's influence. Kantar Shopper Intelligence, 18 September 2026.
https://www.kantar.com/inspiration/brands/kirana-shopkeeper-influence-myth - 02
NielsenIQ (2022). Can the FMCG industry afford to lose billions from empty shelves?
https://nielseniq.com/global/en/insights/education/2022/can-the-fmcg-industry-afford-to-lose-billions-from-empty-shelves - 03
NielsenIQ (2026). India FMCG Quarterly Snapshot, Q2 2026.
https://nielseniq.com/wp-content/uploads/sites/4/2026/08/NIQ-India-FMCG-Quarterly-Snapshot_Q226-AMJ26.pdf - 04
Redseer (2026). Why India's mass grocery still leans on kirana stores.
https://redseer.com/reports/why-indias-mass-grocery-still-leans-on-kirana-stores/ - 05
GS1 Germany and ECR. Category management.
https://www.gs1-germany.de/en/themen/category-management/ - 06
Hennink, M. and Kaiser, B. (2022). Sample sizes for saturation in qualitative research: a systematic review of empirical tests. Social Science and Medicine 292, 114523.
https://pubmed.ncbi.nlm.nih.gov/34785096/ - 07
Gruen, T., Corsten, D. and Bharadwaj, S. (2002). Retail Out of Stocks: A Worldwide Examination of Extent, Causes and Consumer Responses. Grocery Manufacturers of America, FMI and CIES; Corsten, D. and Gruen, T. (2004). Stock outs cause walkouts. Harvard Business Review; FMI (2002). New industry study explores root causes of global out of stock dilemma.
https://www.supplychain247.com/images/pdfs/GMA_2002_Worldwide_OOS_Study.pdf
https://hbr.org/2004/05/stock-outs-cause-walkouts
https://www.fmi.org/newsroom/news-archive/view/2002/10/14/new-industry-study-explores-root-causes-of-global-out-of-stock-dilemma - 08
Knowledge at Wharton (2005). Delving into the mystery of customer satisfaction: a Toyota for the retail market? Quoting Daniel Corsten.
https://knowledge.wharton.upenn.edu/article/delving-into-the-mystery-of-customer-satisfaction-a-toyota-for-the-retail-market/ - 09
McKinsey and Company (2021). The ship has sailed, but many US shoppers won't wait. US Consumer Pulse Survey, 9 to 15 October 2021, n = 2,095; McKinsey (2021). US holiday shopping 2021: Strong demand meets big challenges.
https://www.mckinsey.com/featured-insights/charts/the-ship-has-sailed-but-many-us-shoppers-wont-wait
https://www.mckinsey.com/business-functions/marketing-and-sales/our-insights/us-holiday-shopping-2021-strong-demand-meets-big-challenges - 10
IHL Group (2025). Retail inventory crisis persists despite $172 billion in improvements.
https://www.ihlservices.com/news/analyst-corner/2025/09/retail-inventory-crisis-persists-despite-172-billion-in-improvements - 11
Mint (2025), citing Worldpanel by Numerator. FMCG rides quick commerce for brisk sales.
https://www.magzter.com/en/stories/newspaper/Mint-New-Delhi/FMCG-RIDES-QCOMMERCE-FOR-BRISK-SALES - 12
Bain and Company and Flipkart (2026). How India Shops Online 2026.
https://www.bain.com/insights/how-india-shops-online-2026/
Figures are quoted as published, with their year and base, from the public sources listed above. The paper uses no brand examples; organisations named are cited as sources and are not Nextyn clients unless stated.
Frameworks marked Nextyn analysis are working tools, not measured findings. The worked example is illustrative: the brand, region, figures and findings are invented to show how the workflow runs and do not describe any real company or study.
Hero photograph: Pexels.