River
Y CombinatorBacked by Y Combinator
FREE TEMPLATE

Consulting Knowledge Management Template

Four documents and three sheets that decide which past deliverables are actually safe to reuse, instead of trusting a removed client name.

Free download  ·  No account needed

Asset Register  ·  Sennett Advisory  ·  46 components, 14 engagements  ·  audited 04 Aug

Five rows, one from each ground the library actually clears on

RefComponentClassificationGround it clears onStatus
BG-05Five-level process maturity rubricBackground IPNever one client's dataFreely reusable
WP-02Retail labour model calibration approachWork ProductWritten reuse permission in the MSAFreely reusable
WP-04Five-step process redesign narrativeWork ProductMotivated intruder test, pass cleanReusable, cleared
WP-16Regional headcount benchmark tableWork ProductPassed only once the metro name was dropped and the band widenedReusable, cleared
WP-22Regional comparator benchmark tableWork ProductFails: one of three named comparators in a five-firm marketSingle-client-only

Removing the client's name cleared none of these five on its own. Two cleared on origin and contract, one passed the real test clean, one needed a fix first, and one cannot be fixed at all.

Search this and every result is a storage-and-retrieval comparison: Confluence against Notion against Guru against Tettra, scored on search quality and per-seat price. A widely cited four-layer Notion vault guide ships a templates database tracking name, type and times used. None of them ask what actually matters before a file goes to a second client: is it safe to hand over under the confidentiality clause you signed for the client it came from. That includes the newer tools that draft from a firm's past deliverables.

The folk answer independents give themselves, that removing a client's name counts as anonymizing a file, usually is not true. This pack runs three checks instead. First, origin: a component built independent of any client's data is Background IP and clears immediately. Second, the contract: a written reuse permission, the consulting equivalent of a residuals clause, clears the rest. Third, for everything left: the motivated intruder test asks whether a reasonably competent, determined person could work out whose engagement this was from what remains.

Sennett Advisory, a three-person operations consultancy, has 46 components across 14 engagements. Eighteen are Background IP and 3 more clear on contract, but the remaining 25 were all reused on the old strip-the-name rule. Run through the real test, 12 pass clean, 6 pass only after a fix nobody had made, and 7 fail outright. The identifying detail is the finding itself, not decoration on it. Thirteen of 25 name-stripped reuses, 52 percent, never actually cleared the bar the practice assumed they met. What clears becomes a candidate for the productized service pack.

46 components: 13 of 25 name-stripped reuses did not clear the real test

The Asset Register, Reuse Log, Refresh Schedule, and the Library Standard's four statuses.

Asset Register

One row per component: classification, the ground it clears on, test result, status, and rebuild hours if it has to be rebuilt from scratch.

RefComponentClassificationTest resultStatusHours
BG-01Discovery interview guide shellBackground IPN/AFreely reusable2.0
WP-01Cross-industry benchmarking methodologyWork ProductContract governsFreely reusable5.0
WP-09Onboarding-time reduction caseWork ProductPass cleanReusable, cleared3.5
WP-20Margin improvement caseWork ProductPass, remediatedReusable, cleared5.5
WP-25Unit economics market-sizing modelWork ProductFailSingle-client-only7.5

18 Background IP, 28 Work Product. Of the 28, 3 clear on contract, 12 pass the test clean, 6 pass after a fix, 7 fail outright. 46 components, reconciled every time.

Reuse Log

Every time a component left the library, checked or not, before and after the register existed.

DateComponentChecked first?Outcome
2024-07-02WP-16, senior living benchmarkNoRetested later: needed the metro name dropped
2024-09-30WP-25, market-sizing modelNoRetested later: fails, the figure is the client's own
2025-08-04Baseline: all 46 componentsYes18 clear, 3 contract, 12 clean, 6 fixed, 7 retired
2025-11-05WP-19, vendor concentrationYesCleared status confirmed current, reused
2026-03-02WP-22, regional comparator tableYesChecked again: still fails, market has not grown a fourth firm

Seven reuses went out unchecked before the register existed. Five since have all been checked first. Two of the seven historical ones turn out to have failed a test they were never given.

Refresh Schedule

Every cleared or single-client-only component gets a next-check trigger, not only a fixed date.

RefStatusNext-check triggerNext check
WP-04Reusable, clearedStanding interval2026-08-04
WP-16Reusable, clearedA third senior-living engagement, or the interval2026-08-11
WP-22Single-client-onlyA fourth comparator enters the regional marketWatch
WP-27Single-client-onlyA second comparable acquisition case completesWatch

A pass is not permanent and a fail is not forever. WP-16's status depends on how many other senior-living clients exist, not on the date it first cleared.

Library Standard: the four statuses

Only a completed test moves a component out of not yet tested. Nothing is presumed safe by default.

StatusCountWhat it means
Freely reusable21Background IP (18) or contract-permitted (3). No further test needed
Reusable, cleared18Passed the motivated intruder test, clean (12) or after a fix (6)
Single-client-only7Failed. Stays out of anything new until the population changes
Not yet tested0The default the moment anything new is logged, and treated as blocked

21 plus 18 plus 7 is 46. Every component sits in exactly one of four statuses, and none of them is a guess.

What's in the pack

01

Asset Register

Every component classified Background IP or Work Product, with the ground it clears on and its rebuild-hour cost.

02

Reuse Log

Every reuse logged with its destination and whether the status was actually checked first, not assumed from memory.

03

Refresh Schedule

A next-check trigger per cleared component, since anonymity depends on the client roster, not the date it first cleared.

04

How a Component Gets Classified

The three-step order: origin, then the contract, then the anonymization test, run once per component rather than assumed.

05

Client-safe Anonymization Note

What the practice means by anonymized, and why a component covered by a client restriction still needs its own test.

06

Library Standard

The four statuses a component can carry, who can move it between them, and the cadence for re-testing.

07

Reuse Guidance

How to pull a cleared component for a live engagement: check status, match the population, generalize on the way out, and log it the same way the practice dashboard logs every other hour.

08

Removing the Name Is Not Anonymizing the Deliverable

The standing rule: a component clears on one of three grounds, and a removed client name was never one of them.

How to use it

  1. 1

    Open in River, or take it blank

    Open the pack and send past deliverables from any engagement you can still identify, or download the four Word documents and three CSV sheets and run the classification yourself.

  2. 2

    Sort by origin, then check the contract

    Background IP clears immediately. Everything built from one client's actual data gets checked against that engagement's own contract for a written reuse permission before anything else.

  3. 3

    Run the test on what is left

    Strip the obvious identifiers, then ask whether a reasonably competent, determined person could still work out whose engagement it was from what remains, and fix or retire accordingly.

  4. 4

    Log the reuse, schedule the next check

    Every reuse gets a row with its destination and whether the status was checked first. Rebuild hours saved by reuse surface the same way unbilled time does in engagement profitability analysis.

Frequently asked questions

Is this template free?

Yes. The zip is Word documents and CSV sheets, no account and no card. Edit with AI is the optional half: the agent reads your past deliverables, classifies each one, and runs the anonymization test itself. The rest of the packs sit in the template library.

What format are the downloaded files?

Four Word documents and three CSV sheets in one zip. The Asset Register carries the columns that do the work: classification, the ground it clears on, test result, status, and rebuild hours. Excel, Numbers and Google Sheets open them directly.

What does Edit with AI actually do?

It reads whatever past deliverables you send, classifies each one as Background IP or Work Product, and checks the source contract for a reuse permission. For everything left, it runs the motivated intruder test itself and writes the specific identifying detail behind any fix or retirement.

Isn't removing the client's name enough?

No, and that is the whole reason this pack exists. A benchmark table naming a client as one of three comparators in a five-firm market is identifying with the logo removed, because the rest of the description narrows the field to almost nobody. The test covers the whole remaining description, not the one field you deleted.

Why bother tracking rebuild hours if I don't bill by the hour?

Because a fixed fee still has a real cost behind it. A component that fails the test and gets retired is rebuild time you now have to price into the next proposal instead of assuming away. That is the same floor the rate card pack uses for everything else you sell.

Does a pass ever expire?

Yes, and it can also reverse. Anonymity depends on how many similarly described clients exist, so a component safe when the practice had two clients that fit its description can stop being safe with only one left, and the reverse happens as the roster grows. The Refresh Schedule re-tests on that trigger rather than trusting the date it first cleared.

Test it before you reuse it

Take the Word documents and CSV sheets blank, or send River what you have already delivered and get every component classified, tested, and logged.

Edit with AI