We test high-stakes decisions before you execute them.

We stress-test a single high-stakes decision — a price move, a bid, an M&A closing, a restructuring, an AI programme — against the evidence you already own, before you execute it. Every engagement starts with one named decision, one decision owner and one written verdict. We do not sell advice by the hour; we sell the evidence that lets a board commit or hold.

A short signed verdict document: the decision framed, the evidence tested, a confidence score, the recommendation, and the instruments to execute it or the reasons to hold. It is designed to be read by a CFO, an auditor and a board in the same room, without a supporting deck. Redacted specimens are available on request.

Fixed fees anchored to the value at stake and disclosed before we start — never billed by consultant days. On Executive Missions, a portion of the fee can be tied to a benefits ledger reconciled against your finance system on a schedule you set. If the verdict fails our own nine-gate defensibility test, you do not pay the mission fee. The fee sheet is available on request.

CEOs, CFOs and boards of EUR 30–500M companies facing a decision that must survive audit, financing scrutiny, or a change of shareholder. We work with family-owned groups, founder-led companies, PE portfolio companies and country subsidiaries across Francophone Europe and Francophone Africa and their corridors.

Our refusal list, published in full on the About page: no recommendation without a testable hypothesis; no engagement without a named decision and decision owner; no dashboard without a decision it improves; no AI initiative without a workflow experiment and governance; no value claim without baseline, attribution, owner and measurement window; no pilot without a scale threshold and stop rules. If a mission cannot meet the six, we do not take it.

STREDGELAB is our own production system: it runs the deterministic engines, holds the evidence record and versions every method statement. It is not software we license to you. The engagement itself is a live, human-run method — our people apply the engine to your decision, and you receive the evidence and the verdict, never a seat in a tool. The engines named on the Services pages — MarginLeak X-Ray™, Deal Reality Assurance Twin™, and the rest — are versioned internally, and each carries a written method statement. We share the method statement for the engine your decision would use, on request, before we sign an engagement letter.

Data you send us through the contact form and any documents shared for an engagement sit on managed cloud infrastructure in our declared European production region, encrypted in transit (TLS 1.2+) and at rest (AES-256), with Row-Level Security scoping every read and write. Access is limited to named partners under MFA. We do not train third-party AI models on client data and use no advertising or session-replay subprocessors. Full posture, subprocessor list and incident contact on the security page; lawful basis and retention on the privacy page.

You take one decision on your desk through nine gates across three domains — value leakage, AI readiness, decision quality — on the page, in about two minutes. Each domain is read and reported on its own: there is no total, no percentage and no grade. It ends in a printable memo you keep. It is free, needs no account, and nothing is booked. If you want a written verdict on the decision itself, that is the Decision Sprint (L1).