Carbon footprint for a retailer: purchased goods and freight are the whole picture

The desk where most of a retailer's carbon footprint gets decided has no meter and no chimney. It is the buying desk, and the decision is which supplier, which product, which season, moved how.

Everything on the shelves was grown, made and shipped before the retailer ever touched it, and in the GHG Protocol's accounting those embodied emissions belong to the retailer regardless: scope 3, category 1, goods purchased for resale. For nearly every distribution business that one category, plus the freight that moves it, is the footprint. Measure only what you operate, the stores, the warehouses, the delivery fleet, and you will produce a number that looks complete and misses most of the truth.

Segment the assortment before collecting anything

The paralysis specific to retail is assortment size: thousands to hundreds of thousands of references, none individually significant. Segmentation is the way through, and it has to be decided before collection starts rather than discovered halfway.

Split the assortment into a small number of category blocks, aligned with how purchasing is already organized, and ask one question of each: does a physical or supplier-specific route exist at reasonable cost? Food, textiles and electronics have workable average factors per product family, food is covered in real depth on the French side, and large suppliers increasingly publish their own figures. Those blocks take the better route. The long tail, homeware, accessories, seasonal goods, stays on monetary factors applied to purchase spend, exactly as the physical-versus-monetary rule prescribes.

What comes out is a category-level footprint: imprecise at the reference level, reliable at the level where decisions are actually taken.

Both directions of freight are yours

Upstream, the freight from suppliers and the import flows you pay for are your scope 3, measured in tonne-kilometers by mode. The mode mix is the whole story here, because a small share of air freight can outweigh a large share of sea freight. That makes your air-freight exceptions list, usually driven by stockouts and season launches, a carbon document as much as a cost one.

Downstream has grown with e-commerce: home delivery, whether you run the fleet (scope 1) or buy the service (scope 3), plus the packaging around every parcel. Between warehouse and doorstep the last kilometers dominate, and the levers there, delivery density, failed-delivery rates, pickup points, are the ones the logistics team already optimizes for cost per drop.

Good to know: returns are freight twice over, plus the share of returned goods that never gets resold. Where return rates run high, fashion e-commerce above all, the returns flow deserves its own line in the model. It is one of the few carbon lines a commercial policy can move inside a quarter.

Stores and warehouses: the visible minority

Store lighting, heating and cooling, warehouse energy and your own vehicles are scopes 1 and 2, measured cleanly from invoices and fuel cards, site by site. This is the part management sees every day, and for most non-food retail it is a modest share of an honest total. Put both numbers side by side in the report, because the proportion is the finding, and it is the sentence that stops a year of effort going into store lighting.

Food retail earns one exception, and it is a large one: refrigeration. Commercial refrigerant circuits leak, the recharge records are the measurement, and older high-GWP (global warming potential) refrigerants make every kilogram lost expensive in CO2-equivalent. A grocer's operational footprint deserves a refrigeration block of its own, energy plus leakage, next to the building energy.

From footprint to buying decisions

The category ranking the footprint produces is also your engagement plan. The top categories name the suppliers whose data would improve the model most and whose practices weigh most, and those relationships, a handful at a time, get the supplier questions, the data clauses at contract renewal and the product-level requests. It is the same campaign discipline that applies to component suppliers in electronics. The tail stays on monetary factors and gets re-ranked once a year.

So build the model the way purchasing thinks, not the way the reporting template asks. A retailer's real carbon instrument is the assortment itself, what gets listed, from whom, and moved how, and a footprint segmented by category is the only version that lands on the desk where those three things are decided.

FAQ

What is the largest emission source for a retailer?

Goods purchased for resale, scope 3 category 1, plus the freight that moves them. The embodied emissions of the assortment usually outweigh store and warehouse energy many times over, with food retail's refrigeration as the notable operational exception.

How do you measure emissions across thousands of product lines?

By segmenting the assortment into category blocks aligned with purchasing. Heavyweight categories with usable data, food, textiles, electronics, and large suppliers' published figures, get physical or supplier factors; the long tail stays on monetary factors applied to spend. The result is reliable at category level, which is where buying decisions are made.

Do customer deliveries count in a retailer's carbon footprint?

Yes. Deliveries made with your own fleet are scope 1; purchased delivery services are scope 3. The last kilometers dominate delivery emissions, and returns add a freight flow plus the share of returned goods never resold, which is worth isolating in categories with high return rates.