Results, with the work shown.
Each case walks the full 4D arc, what was diagnosed, what was decided, what was built, and what moved. Numbers included, hype excluded.
Halving churn in a subscription business
A subscription company bleeding customers faster than it could buy them. Retention economics rebuilt from first principles.
Read the case → Client Engagement · AnonymizedUnlocking incremental e-commerce revenue
A multi-channel retailer treating digital as a side project. Channel strategy that made it a growth engine.
Read the case → Client Engagement · AnonymizedExpansion that paid for itself
New-market entry sequenced by evidence instead of ambition, $18M+ in new revenue across five-plus markets.
Read the case → Client Engagement · AnonymizedMargin work that didn't gut the business
Profitability restored through pricing, mix, and operational focus, not through cuts that mortgage the future.
Read the case → Composite CaseAgentic AI where it pays for itself
From seventeen AI subscriptions and no results to three production workflows with measured ROI.
Read the case → Composite CaseA $3M agency chooses, finally
A founder saying yes to everything learns the economics of focus. Conversion +12%, ARR +15% within two quarters.
Read the case →About these cases: Engagements marked "Client Engagement" describe real work with details anonymized to protect client confidentiality; no former-firm clients are identified. Cases marked "Composite" are illustrative composites built from recurring patterns across multiple real situations, and are labeled as such wherever they appear.
Your business could be the next case
Every one of these started with a 30-minute conversation.
Every case is labelled, and the labels mean something.
A consulting site with unverifiable claims is a consulting site that has told you exactly how much it values precision. These are labelled so you can weight them yourself.
Client Engagement · Anonymized. A real engagement. Client identity, sector detail and some absolute figures are withheld or adjusted under confidentiality obligations. The mechanism described (what was diagnosed, what was decided, what changed) is accurate.
Composite Case · Illustrative. A synthesis of recurring patterns across several real situations, not a single client. Written as a narrative because that is how the pattern is legible, labelled as composite because presenting it as one client would be a small dishonesty that undermines everything else here.
Every case also carries a limitations section stating what it does not prove. That section exists because a case study without one is marketing, and because the conditions under which a result was achieved matter more than the result when you are deciding whether it applies to you.
What the cases have in common
The presenting problem was not the real one
In five of six, the question that arrived is not the question the data supported. A channel assumed weak was starved. A cost problem was a mix problem. A growth question was a subtraction question.
Something was deliberately stopped
Seven markets declined. Six service lines cut. Twelve AI subscriptions consolidated. The subtraction is not a side effect of the strategy, in each case it is most of it.
Measurement was designed before the intervention
Baselines captured first, so the result could be demonstrated rather than asserted. Retrofitting a baseline after the fact is how programmes end up unable to prove anything.
About the work
Why are the clients not named?
Because confidentiality obligations from prior clients and employers are absolute, and honouring them without exception is itself the relevant signal. A consultant who names one client's details to win the next engagement will name yours. If that makes a claim harder to verify, that is the correct trade, and the same discipline will apply to your information.
Can I speak to a reference?
Where a client has agreed to it, yes, asked for and arranged at the proposal stage rather than published on a website. That is also the honest position for a young practice: the reference list is short, and pretending otherwise would be the first thing you found out was untrue.
Do results like these transfer to my company?
Partially and conditionally, which is why each case carries a limitations section naming its preconditions. The transferable part is the method, not the number. Any advisor presenting a past percentage as a forecast for your business is either not thinking carefully or hoping you are not.
The interesting case is yours.
Thirty minutes establishing what is actually binding in your business, before anyone proposes a solution shaped like someone else's.