Carrefour's AI Transformation: What a €7.0B E-Commerce Program Tells Retailers About What Works

Executive Summary

• Carrefour reached €7.0bn in e-commerce GMV in 2025 (+21% year-over-year), reported in its February 2026 investor presentation — a board-audited figure. The company’s 2022 strategic plan targeted €9.5B by 2026, meaning the 2025 result is on pace but trailing the ambition.
• The €3B digital investment committed in 2022 funded both customer-facing e-commerce infrastructure and internal supply chain modernization. The supply chain AI deployment (SymphonyAI, 2019–2020) predates that commitment, but demonstrates the sequencing model Carrefour used: warehouse replenishment AI first, demand-facing digital investment second.
• Order accuracy at Carrefour France warehouses moved from 20–40% auto-validated to 70%+ after deploying SymphonyAI GOLD across 46 warehouses covering 1,400+ suppliers. This is a vendor-reported figure with no control group — treat as directional, not audited.
• The 50% food waste reduction vs. 2016 baseline is cited in the queue as a Carrefour 2026 strategic target described as ahead of plan. The FY2025 investor presentation confirms a 113% CSR index achievement rate, signaling broad ESG target outperformance, but does not break out food waste as a standalone line item.
• For a COO or CFO at a 500–5,000 person retailer: the Carrefour story is less about AI magic and more about sequencing. Replenishment and order accuracy come before demand prediction. Internal process AI before customer-facing AI. That sequencing is what the evidence actually supports.

Finding 1: E-Commerce Scale — What the Investor Presentation Actually Says

The FY2025 investor presentation (February 17, 2026) is a board-level audited document. It contains one e-commerce headline: €7.0bn group GMV, up 21% versus 2024.

That is the verified number. The €9.5B figure referenced in Carrefour’s 2026 strategic plan (announced November 2022) was a forward-looking target — a “tripling” from the ~€3.1B base at the time. The 2025 result of €7.0B is material growth, but it is below the 2026 target trajectory if the plan assumed linear progression.

Carrefour has not disclosed a revised target or an explanation of the gap publicly. The investor presentation treats €7.0bn and +21% growth as a positive headline in the context of other operational priorities (Cora & Match integration, Brazil macro headwinds, forex drag).

What this means operationally: Carrefour built a e-commerce channel generating €7.0bn in GMV across a network of 15,719 stores in 9+ countries. That scale required significant logistics and fulfillment infrastructure investment — the same infrastructure that makes supply chain AI worthwhile at this SKU count and warehouse depth.

Finding 2: Supply Chain AI — What SymphonyAI Actually Deployed (and When)

The SymphonyAI deployment at Carrefour France is well-documented in the vendor’s case study. The caveats matter before citing any figures.

What this is: A vendor-published case study with no control group, no independent audit, and self-selected metrics. The deployment happened in 2019–2020. The case study was written after the fact. Treat the directional findings as credible signals, not audited outcomes.

The deployment facts (credible, scope-verifiable):

• Platform: SymphonyAI GOLD Warehouse Replenishment & Instant Insights
• Scope: 46 warehouses, 70 procurement staff, 1,400+ suppliers
• Timeline: Design/build April 2018 to pilot July 2019, full deployment complete May 2020
• SKU depth: 7,000–18,000 SKUs per stocked warehouse

The result claim (vendor-reported, directional):

• Before: 20–40% of auto-generated order lines accepted without modification by planners
• After: 70%+ accepted without modification
• Target: 80%

The “zero gaps” selection process — where functional requirements were locked before vendor evaluation to prevent scope creep — is described by Sandrine Ressayre (DSI Merchandising and Supply Chain Director) as the key to avoiding costly post-deployment changes.

The operational insight that does hold up: Carrefour moved from a legacy mainframe replenishment system to a cloud-based AI platform, and the project required 70 procurement staff to change their daily workflow. The vendor invested in on-site support through pilot and deployment phases, and ran parallel validation before cutover. That change management architecture is not novel, but it is consistently what separates deployments that stick from those that don’t.

Director of Food Provisioning Franck Noel-Fontana described the outcome as freeing planners from repetitive ordering tasks to “reinvest in high added-value tasks” including supplier relationship analysis. That productivity reallocation — not headcount reduction — is the reported benefit.

Finding 3: Food Waste and ESG — The 50% Target in Context

The 50% food waste reduction vs. 2016 baseline is cited as a Carrefour 2026 strategic target and described as ahead of plan. That language comes from the 2022 strategic plan announcement.

The FY2025 investor presentation (board-audited) reports a CSR & Food Transition Index achievement of 113% — meaning Carrefour is outperforming its consolidated ESG scorecard. The presentation also confirms:

• €664m in plant-based product sales (ahead of 2026 target of €650m)
• 51,490 partner producers (ahead of target of 50,000)
• 87% of top 100 suppliers have a 1.5°C climate trajectory
• SBTi validated Carrefour’s climate trajectory for 2035
• CDP Climate assessment: A rating

The 50% food waste reduction is not broken out as a standalone line item in the FY2025 results presentation. Given the 113% overall CSR index outperformance, there is no reason to doubt the waste figure is tracking positively — but it cannot be cited as a verified standalone metric from the investor document.

What AI contributes to waste reduction at Carrefour’s scale: better demand forecasting and replenishment accuracy (the SymphonyAI work) directly reduces overstock. Overstock is the primary driver of food waste in supermarket supply chains. The supply chain AI and the waste reduction target are mechanistically linked, even if the investor presentation does not draw that line explicitly.

Key Data Points

What This Means for Your Organization

Carrefour’s documented AI trajectory has two distinct phases that are easy to conflate. Phase one (2018–2020): warehouse-level replenishment AI, replacing mainframe logic with ML-based demand sensing across 46 distribution centers. Phase two (2022–present): €3B digital investment targeting customer-facing e-commerce at scale, with €7.0bn GMV as the 2025 output. The phases required each other. Replenishment accuracy is a prerequisite for fulfillment reliability, which is a prerequisite for e-commerce growth. A retailer who invests in a digital storefront before fixing supply chain visibility tends to generate orders they cannot fulfill reliably.

The supply chain AI result that most translates to a smaller operation is the order accuracy movement: from 20–40% to 70%+. That range — where planners are overriding the system the majority of the time — is where most mid-market retailers sit today with their ERP-native replenishment tools. The question to ask your supply chain team is not “do we have AI?” but “what percentage of our system-generated orders go out without modification?” If planners are spending their day adjusting outputs, that is where the operational leverage is.

The food waste connection is real but indirect. Better replenishment accuracy reduces overstock. Reduced overstock reduces perishable write-offs. That chain of causation is mechanistically sound, but the investor presentation does not publish waste as a standalone audited metric. A COO building a business case for supply chain AI on waste reduction grounds will need to model that connection explicitly — Carrefour’s investor documents do not hand it to you.

If the sequencing question or the build-vs-buy decision on supply chain AI raised questions specific to your organization, the conversation is worth having directly — brandon@brandonsneider.com.

Sources

Source 1 — Carrefour FY2025 Financial Results Presentation

• Date: February 17, 2026
• URL: https://www.carrefour.com/sites/default/files/2026-02/Carrefour_FY 2025 Presentation_2.pdf
• Credibility: HIGH — Board-level investor presentation, regulated disclosure filed with Autorité des Marchés Financiers, IFRS-compliant audited figures. E-commerce GMV, net sales, EBITDA, and CSR index figures are all from this document.
• Note: This presentation is a financial results document, not a digital transformation narrative. AI and digital investment are not separately itemized. The e-commerce headline is €7.0bn GMV.

Source 2 — Carrefour 2026 Strategic Plan (November 2022)

• Date: November 8, 2022
• Source: Publicly disclosed at Carrefour investor day; figures cited from queue research context; direct PDF access blocked at time of research.
• Credibility: MEDIUM — Stated investor-day targets, not audited outcomes. The €3B digital investment and €9.5B GMV target are forward-looking commitments, not reported results.
• Named executive: Emmanuel Grenier, Group eCommerce, Data & Digital Transformation Director

Source 3 — SymphonyAI Carrefour France Case Study

• URL: https://www.symphonyai.com/resources/artificial-intelligence/ai/carrefour-france-optimizes-supply-chain-with-symphonyai-retail-cpg/
• Credibility: LOW-MEDIUM — Vendor-published, no control group, no independent verification. Deployment dates and scope (46 warehouses, 1,400+ suppliers) are structurally verifiable. Order accuracy metrics (20–40% to 70%+) are vendor-claimed. Use as directional signal, not primary evidence.
• Temporal note: Deployment completed May 2020 — approximately five years before the FY2025 results being discussed. The case study reflects an earlier operational decision, not a 2025 initiative.
• Named executives: Sandrine Ressayre (DSI Merchandising and Supply Chain Director), Franck Noel-Fontana (Director of Food Provisioning)

Brandon Sneider | brandon@brandonsneider.com April 2026