How Picnic modeled a 24% water footprint cut before changing a single supplier

The scope 3 water footprint problem in retail

A typical retail assortment is made of thousands of products, each built from many materials, each produced across multiple locations. For nature assessments, those sourcing locations matter, and that hierarchy of products, materials, and production steps can explode into millions of data points.

Carbon has one metric and one target: global warming potential. Nature doesn’t work that way. Sustainability teams have to track up to 15 different environmental indicators, most teams aren’t yet trained on the topic, and almost all of the impact sits in scope 3, outside a company’s direct control. Change one thing to improve your carbon footprint, and you can quietly make your water footprint or land footprint worse. Understanding that trade-off before making a buying decision requires scenario modeling, not a dashboard.

“

“The platform automatically generates these highlights and hotspots, making everything very clear right from this view.”

Picnic’s supply chain, ten years in

Picnic is an online supermarket operating in the Netherlands, Germany, and France, built less like a traditional store and more like a modern milkman, with electric vans delivering direct from a dedicated supply chain. A decade in, Picnic operates in more than 700 cities with over 20,000 employees and a private-label range of 3,000+ SKUs, the layer of their assortment where they have the most direct influence.

Like most supply-chain-heavy retailers, almost all of Picnic’s footprint sits in scope 3, with a large share tied to the food they sell. Picnic has financing linked to scope 3 progress and a growing list of initiatives to move it, but each initiative changes their risk profile differently. Shifting toward more plant-based products affects their footprint very differently than changing where an ingredient is sourced. To act on scope 3 with confidence, Picnic needed to see both paths before choosing one, which is what brought them to Spirefly.

Finding the water footprint hotspot: rice sourced from Pakistan

Working in the Spirefly Workbench, Picnic’s team uploaded their assortment data and let the platform surface hotspots automatically. On the carbon side, the answer was familiar: two suppliers, one meat, one dairy, stood out immediately, with milk, beef, other meats, nuts, and rice as the ingredients driving the impact.

The more useful discovery came from going beyond carbon into a full nature assessment. Layering in water and land use shifted the focus from what to where. The Netherlands showed up as expected, as a major dairy source. But Pakistan emerged as an outsized water user for the assortment, tied closely to the basmati rice Picnic sources from there. That combination, one ingredient, one country, one disproportionate water footprint, wasn’t visible from carbon data alone, and it gave the team a specific, testable question: could that rice be sourced from somewhere less water-stressed?

Modeling a 24% water footprint reduction with scenario planning

That’s where Spirefly’s scenario builder came in. Picnic modeled what would happen if they shifted their basmati rice sourcing from Pakistan to Italy, without contacting a single supplier first. The result: an estimated 24% reduction in the water footprint for that part of the assortment. The same exercise applied to soy and palm oil sourcing surfaced comparable reductions in other high-impact categories.

The team ran the same kind of test on the demand side. Modeling a shift from cow’s milk to a plant-based oat variant showed a 6% reduction in land use, and because scenarios can be projected forward in time, Picnic could also model category growth (say, 5% more milk sold per year) alongside the substitution, to see how the two forces net out by 2030.

A third scenario looked at one of Picnic’s meat suppliers directly: what would deforestation-free feed do to that supplier’s carbon emissions, versus a switch to renewable energy at the processing stage? The feed change moved the needle meaningfully; the renewable energy switch barely did, a useful, counterintuitive finding, since most of that supplier’s footprint traces back to how the cattle are raised and fed, not how the meat is processed.

Turning biodiversity risk into buyer decisions

None of this becomes useful until it reaches the people who make buying decisions. At Picnic, that means the sustainability team doesn’t approach a supplier directly. They take a hotspot, like the rice from Pakistan, to the category manager who owns that supplier relationship, translate the finding into buying language, and let the category manager decide whether and how to start the conversation. From there, it becomes one input among several in the trading team’s decision-making, alongside price, quality, and supply reliability.

Key takeaways for scope 3 water footprint reduction

Start with the data you have, not the data you wish you had. Spend-based estimates and partial ingredient data are enough to get a first, prioritized view of risk. Just be honest that more assumptions mean less certainty, and treat early hotspots as a place to start asking questions, not a final verdict.

Model before you commit. Scenario modeling lets a sustainability team explore sourcing changes, product substitutions, and supplier interventions on paper, understand which ones move the metric that matters without shifting the burden elsewhere, and bring a category manager a specific, tested idea instead of an abstract concern.

Translate risk into the language buyers already use. A biodiversity risk only drives change once it shows up as one more factor in a category manager’s existing decision-making process, alongside the criteria they already weigh.