Tracking One Category Across Four Countries: Butter Brands and a Weekly Dashboard

The Problem: Cross-Border Monitoring Usually Gets Scoped Too Wide

A brand selling into several export markets wants to know how it is priced against competitors in each. The instinctive scope is broad — monitor the grocery retailers in each market, capture everything, analyse later.

That scope fails in a specific way. Every market has different retailers, different catalogue structures, different languages and different category taxonomies. Multiply "capture everything" by four countries and the engineering effort goes into breadth the brand will never look at, while the questions it actually has — how is Lurpak priced against us in this market, this week — stay unanswered because the data is too broad to be maintained reliably.

This client inverted it. One category. Named brands. Named retailers per market. Weekly.

The brief was to monitor pricing and availability of butter products across selected e-commerce supermarkets in multiple countries, tracking a specific competitive set — Lurpak, President, Blue River, Kerrygold — from defined target retailers including Riba Smith and Rey Supermercados, with Panama added as a new market during the engagement.

Why Narrow-and-Deep Is the Right Shape Here

• Named brands make matching tractable. Cross-border product matching is the hard part of multi-market comparison — different languages, different pack conventions, different local naming. Restricting to four named brands reduces this from an open-ended entity-resolution problem to a bounded one. You are not matching a catalogue; you are finding four brands' products and normalising their pack sizes.
• One category means one taxonomy problem. Butter sits in a different place in each retailer's category tree, and finding it once per retailer is a solvable setup task. Doing that for every category in every retailer is a permanent maintenance burden.
• Weekly matches the actual price rhythm. Grocery base prices and promotional cycles move weekly. Daily collection on this category would re-read unchanged data six times a week.
• Adding a market becomes configuration, not construction. This is the payoff and it was demonstrated in the engagement. Panama was added as a new market — new retailers, same category, same brand set, same schema. Because the scope was narrow and the schema stable, onboarding a market meant defining retailer-specific category paths, not designing a new pipeline.

What Made It Work

• Scope by category and brand, not by retailer catalogue. This is the transferable lesson. Cross-border monitoring becomes affordable and maintainable when the scope is defined by what you actually compete on.
• Normalised unit price as a first-class field. Without it, cross-border comparison is not possible at all. With it, market entry pricing, export positioning and promotional response all become measurable.
• Ship a dashboard when the audience isn't analysts. The delivery format determines whether data gets used. Commercial teams need a view; analysts need the file. Providing both costs little and roughly doubles the internal audience.
• Design so new markets are configuration. The Panama addition proved the architecture. If adding a market requires engineering, the scope was wrong.

Where This Applies

The pattern suits any brand competing in multiple markets on a defined category: FMCG and packaged foods, dairy, beverages, personal care, home care, pet food, and private-label benchmarking against branded competitors.

Our grocery and retail footprint spans roughly 148 distinct platforms across India, the US, UK, Australia, Malaysia, Lebanon, New Zealand, Panama, Nigeria and Singapore, including Aldi, Woolworths, Coles, Wegmans, Sainsbury's, Sam's Club, Costco, Metro Cash & Carry, BigBasket, DMart, JioMart and Udaan.

Typical entry scope: one category, three to five named competitor brands, two markets, weekly, for four weeks. It establishes normalisation quality and the real size of your cross-market price gaps before wider commitment.

FAQ

How do you compare grocery prices across countries with different pack sizes?

Through a normalised per-unit price computed from the pack size and unit on each listing, held alongside the original pack information and local currency. Absolute prices are not comparable across markets; normalised unit prices are.

Is it better to track a whole catalogue or one category?

For competitive benchmarking, one category with a named competitor set is almost always better. It is cheaper, more maintainable, and produces the specific answers commercial teams need. Full-catalogue collection suits assortment and market-structure research, not price positioning.

Can data be delivered as a dashboard rather than files?

Both. A Power BI dashboard was delivered alongside weekly Excel/CSV files in this engagement. Dashboards serve commercial teams; files serve analysts and preserve the raw asset.

How quickly can a new country be added?

Where the schema and category scope are stable, adding a market is a configuration task — identifying the target retailers and their category paths. Panama was added this way during the engagement rather than as a new project.

How often should grocery category pricing be collected?

Weekly suits grocery base prices and promotional cycles, which move on a weekly rhythm. Higher frequency mostly re-reads unchanged data unless quick commerce is in scope.

What happens when a tracked brand disappears from a retailer?

It is flagged, then investigated as either a genuine delisting — which is a valuable competitive finding — or a collection failure. Treating the two as the same thing is how monitoring programmes lose credibility.