River
Y CombinatorBacked by Y Combinator
FREE TEMPLATE

Consulting Findings Report Template

Five documents and four sheets built so every finding names the standard it is measured against and every recommendation names the person who says yes.

Free download  ·  No account needed

Finding Register  ·  F-02  ·  governance review, 2,400 home housing association

One row, and the first field is the whole argument

FieldWhat the register holds
StatementRepairs are recorded as complete before the tenant confirms the work
CriteriaThe repairs policy defines completion as the tenant confirming the work is done
Repairs policy section 4.2
Condition31 of 60 sampled jobs were closed on the operative's return to the depot, with the tenant confirmation field empty
CauseThe mobile app marks a job complete when the operative taps done, and the confirmation field is optional
EffectReported first-time-fix performance is overstated by an unknown margin, and the 14 percent recall rate is the only visible symptom
PerspectiveSystemic  ·  31 of 60  ·  sampled from 2,410
Repeat ofF-04 in the 2024 review
Officials' viewHead of repairs agrees the field is optional and states the policy definition was never implemented in the app

Drop the criteria row and this becomes an opinion. Repairs reporting is inconsistent has nothing in it to agree with, disagree with or act on. A named policy section and a counted sample has one question left in it, and the question is what they intend to do.

Repeat of F-04 is the sentence that changes the room. A new finding invites a discussion about whether it matters. The same finding marked as a repeat of a 2024 one, with the reference, invites a discussion about why nothing happened, and that discussion belongs to the client.

31 of 60 sampled from 2,410. The register never says several, and it never extrapolates a convenience sample to the population.

Search this and you get a table of contents. Title page, executive summary, background, methodology, findings, recommendations, appendices. Every result on page one specifies the order of the sections and not one of them specifies what has to be inside a finding. So the section gets filled with sentences like repairs reporting is inconsistent, which is an observation with a judgement attached, and the client thanks you for a thorough piece of work and decides nothing. The layout was never the problem.

A finding is a structure. 2 CFR 200.516(b) sets out what a federal audit finding has to contain, and it is the most widely applied published specification of one. Five elements: criteria, condition, cause, effect, and whether the case is isolated or systemic with the count stated against the number examined. Criteria is the element consulting drops. The policy defines completion as the tenant confirming the work. Thirty-one of sixty sampled jobs closed at the depot with that field empty.

The decision side follows 2 CFR 200.521, which requires a management decision on each finding stating whether it is sustained, the reasons, and the expected action, carrying the finding's own reference number. So every recommendation traces to a cause and arrives as a question with a yes and a no, a named role and a date. Build it from what an interview programme and a scored assessment produced, or take the files blank.

The row with no criteria, and the refusal that counts as a result

Every recommendation quotes the cause it addresses. Every decision carries the finding reference, so in a year the decision can still be read back against what it was made about.

Finding Register

Extract for a fictional housing association, Brackenfield. Seven rows, one of which is not a finding.

RefCriteria and its sourceConditionPerspectiveSev
F-01Management agreement clause 8.3 requires a monthly board performance reportReports exist January to March, nothing April to September, no board minute records the gap6 of 6 monthsHigh
F-02Repairs policy section 4.2 defines completion as tenant confirmation31 of 60 sampled jobs closed at the depot with the confirmation field empty31 of 60High
F-03Standing orders paragraph 12 requires every variation signed by both parties before work proceedsTwo variations agreed by email, neither executed, work proceeded and was invoiced2 of 4Medium
F-04Performance framework appendix B requires a documented calculation per published indicatorThree staff produced three different figures for the same quarter, published figure matches none3 of 3High
F-06No criteria identified  ·  searched six documentsThe rota names a duty manager nightly. Where that person is unavailable, practice variesNot countedObservation
F-07Stores procedure table 3 sets a minimum holding for 22 emergency parts4 of 12 spot checks found a part below minimum, twice the same part4 of 12Low

F-06 is the row that makes the other six believed. Three people described three different informal cover arrangements, which is genuinely worth telling the client. Nothing in the rota procedure, the job descriptions or the policy requires cover, so it is an observation and the report says so in one line. Inventing the criteria from professional judgement is what produces the readout where somebody asks who says we have to do that.

Severity is set against the effect, not the instance count. F-03 is two contracts of four and it is medium because the money is recoverable. F-04 is three reproductions of three and it is high because the figure is published externally.

Views of the responsible officials are gathered before the report is issued and written into the row, including the two rows where the officer disagrees.

Finding to Evidence Map

One row per element per finding. A partly supported element is a filled row, not a blank one.

RefElementKindWhat it isHolds?
M-06F-02 conditionRecordJob level export, 60 jobs, closure timestamp and confirmation fieldYes
M-07F-02 causeObservationWatched an operative close a job with the confirmation field untouchedYes
M-03F-01 causeStatementTwo directors independently describe the post being backfilled without the taskYes
M-04F-01 effectAnalysisThe two service decisions taken in the window and what each citesPartly
M-12F-04 causeStatementPerformance lead attributes the lost definition to the 2022 migrationPartly
M-15F-06 criteriaNoneSearched the rota procedure, the job descriptions and the policyNo

M-07 is the strongest row on the page. The cause is a configuration choice and an observation proves it rather than describing it. Compare M-12, where a single interested witness offers the origin of a lost definition: the finding records it as the origin offered, not as established cause, and the recommendation says what would establish it.

M-04 is a partly that matters. The minutes show no performance basis for either decision. Whether the decisions would have differed is not established, so the finding does not claim it. Overclaiming an effect is how a correct finding gets thrown out on the one sentence that overreached.

M-03 is a statement carrying a cause, which is allowed here because two accounts agree independently and neither witness has an interest in the version given.

Recommendation Register

Every row quotes the cause it addresses. A recommendation aimed at a condition is a training reminder.

RefFindingRecommendationCause it addressesDecidesBy
R-02F-02Make the tenant confirmation field mandatory before a job can closeThe field is optional in the configurationExec dir housing14 Nov
R-03F-02Restate first-time-fix for the year once confirmed closures are the basisThe figure rests on operative closures, not confirmationsChief executive21 Nov
R-04F-03Name one post accountable for contract administrationNo single owner after the procurement post merged into financeFinance director14 Nov
R-06F-04Publish one documented calculation per indicator and reconcile the three figuresThe definition was lost with an inaccessible spreadsheetExec dir housing21 Nov
R-10F-07Set an automatic reorder point for the 22 minimum-hold partsReorder is manual and uncovered during leaveDir operations14 Nov

R-02 and R-03 are the same finding and two different jobs. One closes the gap going forward, one deals with the number already reported. Practitioners write the first and skip the second, and the second is the one the regulator asks about.

Notice what is not in the R-02 row. Retrain the operatives on closure procedure. The condition is that jobs close early; the cause is an optional field. A recommendation aimed at the condition expires the week the last trained operative leaves.

Two findings sharing a cause share one recommendation, with both references on it. Ten recommendations across four decision makers is the running order for the readout, grouped by person and cheapest first.

Decision Log

Five outcomes, and silence is not one of them.

RefRecOutcomeSustainedReason recorded
D-02R-02ApprovedYesAccepted, including the consequence that reported first-time-fix will fall
D-03R-03DeferredYesWants the restated figure in hand before deciding who receives it
D-05R-05Not approvedYesCounsel judged unnecessary: both counterparties confirmed the variations in writing and neither is in dispute
D-06R-07SupersededYesOvertaken by R-06 now that the reconciliation is funded
D-08R-06AwaitingYesAsked in the readout and not answered. Two working days past the agreed date
D-09R-08AwaitingDisputedEffect disputed by the customer services manager, who has asked for the regulator return arithmetic first

D-05 is a good outcome. The finding is sustained and the recommendation refused, because the chief executive knows something the review did not: both counterparties confirmed the variations in writing and neither is in dispute. The exposure is accepted, in writing, by the person entitled to accept it. Argue with that refusal once and every later decision maker learns that no costs them a debate, so they start saying nothing instead.

Disputed is a state of the finding, not of the decision. The customer services manager disputes the effect on D-09. That dispute lives in the finding's own views field, the arithmetic went back the same day, and the decision request still gets asked.

Days past the agreed date is the only number in this pack that is supposed to make somebody uncomfortable, and it belongs in the status report rather than in a footnote.

What's in the pack

01

Finding Register

Criteria with its source clause, condition, cause, effect, perspective as instances over cases examined, prior-review reference, and the accountable officer's own recorded view.

02

Finding to Evidence Map

One row per element per finding, four evidence kinds in strength order, and a partly supported element written out rather than left blank.

03

Recommendation Register

Each row quotes the cause it addresses, carries the owner role and effort, and names the decision it needs from whom by when. Where the decision needs money, the value model that prices it against their own account codes is the sheet finance asks for.

04

Decision Log

Approved, not approved with the finding sustained, deferred, superseded or awaiting, with the reason and what changes recorded against the finding reference. Where several interdependent recommendations need a dedicated session, the decision workshop pack checks each one's prerequisite and stakeholder alignment first.

05

What Makes a Finding

The five elements, where criteria legitimately come from, and the rule that demotes a finding with no standard behind it to an observation.

06

Decision Request Format

Seven lines per decision, including what stays the same on a no and the consequence the decision maker has not thought of.

07

Findings Report and Executive Summary

Seven sections with observations separated from findings in one line, and a one page summary that carries the decisions rather than the methodology.

08

Readout Running Order

Ninety minutes ordered for decisions instead of for completeness, with the decisions grouped by person and asked out loud in the room.

How to use it

  1. 1

    Open in River, or take it blank

    Open the pack in River and send the evidence with the client's own contract and policies, or download the Word documents and CSV sheets and fill them yourself.

  2. 2

    Find the criteria first

    Every finding gets a written requirement behind it, cited to the clause, section or paragraph. Anything with none becomes an observation and the report says which is which.

  3. 3

    Trace each element to evidence

    One map row per element. A cause resting on one interested witness is recorded as the origin offered, and the effect never claims more than the evidence carries.

  4. 4

    Ask the decisions in the room

    Grouped by the person who answers, cheapest first, with the date written down out loud. Then read the log back and send it the same day.

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 hunts the criteria inside the client's own contract and policies, then builds the registers from your evidence. The rest sit in the template library.

What format are the downloaded files?

Word (.docx) for the five documents and CSV (.csv) for the four sheets, zipped together. Excel, Numbers and Google Sheets open the finding register, the evidence map, the recommendations and the decision log straight off the download.

Why does a consulting report need audit finding structure?

Because the elements are the same and one profession wrote them down. A federal single audit finding must carry criteria, condition, cause, effect and perspective. Consulting produces the same object, calls it a finding, and routinely ships it without the first element.

What if there is no policy behind something I found?

It becomes an observation, with the criteria field reading none identified and the search you ran recorded. It still goes to the client, in its own section. Inventing the standard from professional judgement is what loses you the room when somebody asks who requires that.

The client disputes one of the findings. Do I soften it?

No. Their view goes into the finding, verbatim where it is short, and the arithmetic stands. A disagreement written into the document is a stated disagreement rather than an ambush in the readout, and the decision request still gets asked. Where the pushback runs across several review rounds, the client feedback and iteration pack logs why each round's request was accepted or declined.

Where do the findings come from?

Whatever the engagement produced: analysis, interview synthesis, system exports, working papers. The criteria come from the client's contract and policies, which is why the intake pack asks for those documents on day one rather than in week three.

How is this different from a maturity assessment?

An assessment grades the organisation against a scale and hands back a profile. This states specific gaps against specific written standards and turns each one into a decision with a name on it. Consultants run both, in that order, and the deliverables are not interchangeable.

Turn evidence into decisions

Take the Word documents and CSV sheets blank, or send River your draft findings with the client's contract and get back the criteria behind each one.

Edit with AI