The Key Account Gambit — Liant Composants' Portfolio Defence

A four-round, advanced B2B Key Account Management simulation set inside Liant Composants, a EUR 72M French industrial-fastening SME near Lyon (28% gross margin, 180 active B2B accounts, just 6 strategic account managers, the top 10 accounts producing 62% of revenue). On 11 June 2026 the anchor account — Aérospatiale Garonne, an aerospace OEM worth 18% of revenue (EUR 13M/yr) — issues an ultimatum: a 9% price cut plus a dedicated on-site engineer, or it dual-sources 40% of its volume to a German rival within the quarter. Playing the new Commercial Director, you must defend the portfolio without bleeding it dry, working a fixed EUR 1.5M commercial budget that cannot grow this year. Round 1 — Read the Portfolio: discover that revenue is a misleading lens, surface the 40 small accounts that lose EUR 1.1M/yr and the 22 under-served high-margin mid-tier accounts (EUR 9.5M), and compute true profitability (margin minus cost-to-serve) once you can see that 55% of KAM time is trapped on the top 3. Round 2 — Allocate Effort & Set Service Levels: reallocate the 6 KAMs across segments and design differentiated service tiers — moving small accounts to self-serve to recover the leak while funding a higher tier for high-potential mid-tier accounts, all inside EUR 1.5M. Round 3 — Choose the AI Tool & Defend the Anchor: deploy ONE of three AI modules (churn prediction, account scoring, next-best-action) on data that needs EUR 90k of cleanup, and answer the 9% ultimatum — a volume-tied discount, a total-cost-of-ownership value reframe, selective concessions, a blunt across-the-board cut, or a graceful partial walk-away keeping the profitable 60%. Round 4 — Recover & Set the Operating Rhythm: report the four KPIs versus the Round 1 baseline and lock in the new commercial model. The math rewards value-and-potential prioritisation over revenue size, differentiated cost-to-serve, discounts that buy a commitment, focused single-module AI deployment, and growing share-of-wallet inside existing relationships — and measurably punishes the five classic errors: revenue-led effort allocation, uniform premium service, caving to a blanket 9% discount for nothing in return, switching on all three AI modules thinly, and chasing new logos while ~35% share-of-wallet sits untouched. Final KPIs track Account Profitability (EUR M), Retention, Share-of-Wallet, and Cost-to-Serve.