Three-line interview script
A version each for business, risk and finance. Questions target data sources and filing nodes, never satisfaction scores.
- Interview notes
- Conflict table
Capability · Strategy & blueprint
Blueprint artefacts have to survive a signature, not a slide deck
One term, three calculations: business, risk and finance each compute it their own way. The blueprint forces those definitions onto one sheet that three functions sign — a break inventory, process swimlanes, data dictionary v1, a phased roadmap and the business case. The first two steps contain no code.
Position
Build before the definitions settle, and code becomes the argument
Assessment starts on three lines: deal and portfolio owners, risk and compliance, finance and disclosure. We interview each line for more than one round and record every case where the same metric produces a different number in a different department. We then map where data stops moving by itself between systems and spreadsheets.
The blueprint is not a report pack. It is four artefacts downstream teams can quote directly: a break inventory, process swimlanes, data dictionary v1, and a phased roadmap with the business case. Delivery starts only after business, risk and finance sign the dictionary and the roadmap.
No code in the first two steps is a rule, not modesty. Once a live workflow carries an unresolved definition, the cost of changing it loses to the cost of keeping it, and two definitions end up coexisting. The filing season pays the difference.Catalpa delivery method
Six steps, in an order that cannot be shuffled
Each step has an artefact and a signing function. Doing step four before step three is the most common rework cause we see.
Interview three lines, inventory the material
At least one round per line. We collect live spreadsheets, supervisory filing templates and exports from systems in use. Output: interview notes and a definition-conflict table.
Map data flow and locate the breaks
Every manual hand-off is recorded: which table the data leaves, who re-keys it, which filing definition it lands in. Breaks rank by filing burden, not by department. Output: break inventory.
Set the target process and swimlanes
Screening, diligence, committee, closing, post-investment and disclosure sit on one swimlane map, with each node marked as system-carried or human-judgement. Output: process swimlanes.
Write data dictionary v1 and name field owners
Source, calculation, update frequency and a named owner per field. Business, risk and finance sign together. Output: data dictionary v1.
Agree the benefit formulas before anything is built
Filing cycle, manual-reconciliation hours, definition-consistency rate and alert-confirmation rate get their formulas in the blueprint, so acceptance recalculates them the same way.
Phase the roadmap and price the business case
Release order follows supervisory deadlines and filing pressure, with the current manual-hours baseline and the reduction each phase should deliver. Output: roadmap and business case, into the work-breakdown structure.
Artefacts, duration, acceptance
Standard ranges; the clock starts when the work-breakdown structure is agreed. Group programmes spanning several business lines are re-sequenced against the roadmap rather than compressed.
| Artefact | Typical duration | Acceptance method |
|---|---|---|
| Interview notes and definition conflicts | Assessment, 2–3 weeks | Each line owner confirms no conflict is missing |
| Data-flow map and break inventory | Assessment, 2–3 weeks | Breaks matched one-to-one to live filing templates |
| Process swimlanes (deal to disclosure) | Blueprint, 3–4 weeks | Business owner signs every approval and committee node |
| Data dictionary v1 and field-owner matrix | Blueprint, 3–4 weeks | Three signatures; no field left without an owner |
| Benefit metric formulas | Blueprint, 3–4 weeks | Numerator and denominator exist in system data; recalculated once on the spot |
| Phased roadmap and business case | 5–7 weeks in total | Milestones, budget bands and baseline figures agreed |
After signing
How a blueprint becomes a checkable result
A signed dictionary is the start, not the end. After go-live we run a quarterly definition review against the same formulas, so "the filing cycle fell" stays a recorded fact rather than a line in a closing report.
- Consistency rate tracked per period
- Findings shared with three functions
- Deviations return to the dictionary
Four benefit definitions and the trap in each
This table matches the acceptance definitions on our client-impact pages. One principle governs it: numerator and denominator both come from objects already in the system, never from a form someone fills in, or the metric decays within six months.
| Definition | Calculation | Common trap |
|---|---|---|
| Filing cycle | Calendar days from template issued to consolidated group submission | Measuring one rehearsal, missing the quarter peak |
| Manual-reconciliation hours | Filing-season overtime hours ÷ people involved (hours captured by system) | Replacing logged data with interview recall |
| Definition-consistency rate | Share of cases where a metric reads identically across three levels | Sampling only head-office platforms, skipping districts |
| Alert confirmation rate | Share of triggered alerts the business confirms as real risk | Treating "no false positives" as "no misses" |
Instruments
Six tools, each with a hand-over format
We do not sell "consulting capability". We hand over files a delivery team can read without asking us: tables, maps, dictionaries, roadmaps.
Flow and break map
Systems, tables and departments on three axes, with every manual hand-off ranked by filing burden.
- Break inventory
- Hours baseline
Target swimlane map
Pre-investment, portfolio, post-investment and disclosure nodes aligned to accountable roles on one sheet.
- Approval nodes
- Role ownership
Data dictionary template
Five settled columns per field: name, source, calculation, refresh frequency, owner.
- Dictionary v1
- Owner matrix
Filing-burden model
Current manual-reconciliation hours converted into a comparable cost line for the business case.
- Cost baseline
- Per-phase range
Phased roadmap
Release order anchored to supervisory deadlines, with a signable milestone per phase.
- Release order
- Milestones
What the client provides during assessment
The fuller this list, the closer the break map is to fact. Missing items are marked as unconfirmed rather than inferred.
Live spreadsheets and ledgersSupervisory filing templatesApproval policy and committee rulesOrganisation chart and role notesExports from systems in useValuation and disclosure schedulesLP / GP report samplesCurrent deployment and integrations
Best fit
Who needs this step first
In provincial, municipal and district hierarchies, and in fund-of-funds structures, conflicts come from levels. In market managers holding multi-currency, multi-share-class portfolios, they come from valuation and exit models. The work-breakdown differs, so we scope and price the two separately.
- Government guidance funds / FOF
- Provincial and municipal SOE platforms
- Financial holding and quasi-financial groups
- Market PE / VC managers
- Family offices
Published method
We put our judgements in writing
Definitions before features, how to read ranking statistics, why implementation schedules compress — these arguments run on our insights page, ready to cite in an internal paper.
Read the articleWho should not buy this
A single-fund GP with fewer than fifteen people: a shared workbook and one part-time definition owner is cheaper, and a blueprint would push process beyond the management span.
Any institution that will buy the standard edition and staff no definition maintenance: the dictionary drifts after signature and manual reconciliation returns within two years, so the fee buys nothing.
Anyone wanting conclusions first and evidence afterwards. We do not write feasibility papers for others to sign; the signatories must be your own three line owners.
Questions before you commission it
These answers match our contract terms, so they can go straight into an internal approval note.
Can we buy the blueprint without the build?
Yes. Assessment and blueprint stand as their own contract. You leave with the break inventory, swimlanes, dictionary v1 and roadmap, and procure delivery elsewhere. Our position does not change: implement before definitions settle and whoever builds will rework.
How much of our time does this take?
One owner per line who can settle definitions — business, risk, finance — plus an IT colleague to produce system exports. Those people sign the conflict table and the dictionary, so their hours cannot be substituted by a status pack.
We already run an in-house system. Still do a blueprint?
Usually the scope narrows to definition review: align existing fields with external filing outlets, add the dictionary and the change process, rather than redraw every flow. That work sits under data governance and metrics.
Is 5–7 weeks a commitment?
It is the standard range, measured from agreement of the work-breakdown structure. Multi-line group programmes follow the phased roadmap; we do not shorten interviews to hit a date.
Will the business case turn into a feasibility essay?
No. It answers two questions only: the current manual-reconciliation baseline, and the reduction each phase should deliver. You agree the formulas during the blueprint, and every figure comes from objects already in the system.
Can we run platform procurement in parallel?
You can, but workflow configuration and the permission matrix wait for the signed dictionary. We suggest signing generic terms before the roadmap closes, so nobody builds against two definitions.
Next step
Start with a 45-minute current-state session
Bring the spreadsheets and filing templates you use today. We mark the breaks on the wall and give a first read on where definitions conflict, with nothing pre-committed.