Advisory engagements

Start with a decision—not a transformation programme.

Each engagement is designed to answer one commercial question with historical evidence, an explicit profit definition, agreed risk guardrails and an implementable policy handoff.

Three ways to begin

Match the work to the decision maturity.

A diagnostic can stand alone. Deeper work only follows when the first analysis proves useful and the institution has an owner for implementation.

Decision Opportunity Scan 1–2 weeks

A focused framing exercise to identify where an independent replay is most likely to be measurable, governable and commercially material.

Best forTeams with several possible use cases but no agreed first decision.
HandoffPrioritised decision map, data checklist and pilot hypothesis.

Historical Profit Replay 4–6 weeks

Reconstruct one policy from historical decisions and outcomes, quantify cohort-level leakage and compare a transparent challenger under fixed guardrails.

Best forPricing, renewal, approval, collections or channel decisions with usable outcome data.
HandoffBaseline, opportunity view, challenger rules and measurement pack.

Champion–Challenger Advisory Retained

Independent oversight while the internal team pilots, monitors and refines the selected policy in a controlled operating environment.

Best forTeams with internal implementation capacity that want an external decision-quality layer.
HandoffReview cadence, drift checks, exception analysis and policy recommendations.

The core diagnostic

What a historical profit replay contains.

The deliverable is structured for CFO, CRO, credit, product and operations review—not just for an analytics audience.

Decision contract

The precise business decision, owner, success metric, guardrails, exclusions and analysis population.

Data lineage

A field-level map from decision inputs and exceptions through to realised repayment, losses, recoveries and operating costs.

Current-policy baseline

Cohort performance showing where the existing rule creates value, where it leaks and which segments require caution.

Challenger comparison

A transparent alternative replayed on the same historical population, with value and guardrail movement made explicit.

Implementation pack

Decision bands, trees or priorities; monitoring metrics; exception rules; governance notes; and a staged test design.

Good pilot conditions

What must be true before the work starts.

A useful result depends more on decision clarity and outcome data than on sophisticated infrastructure.

A good fit

The institution can identify:

  • One recurring decision with a named business owner
  • Historical decision or offer records
  • Repayment, loss, recovery or conversion outcomes
  • A reasonable view of funding and operating costs
  • Non-negotiable risk and policy constraints

Not a good first fit

Pause when:

  • The goal is approval growth without a loss-adjusted profit definition.
  • There is no reliable link between the decision and its eventual outcome.
  • The sponsor expects an immediate production model or platform replacement.
  • The required change has no internal owner or governance route.
  • The data cannot be shared safely even in de-identified form.

Who typically owns the conversation

One decision. Several legitimate lenses.

CFOContribution profit, funding economics and measurable return.
CROLoss, policy, fairness, stability and governance guardrails.
BusinessAdoption, operational capacity, customer impact and implementation.

Unsure which portfolio decision is worth replaying first?

Send me the shortlist →

A useful note is enough: the decision, portfolio, available history and what currently feels wrong.