INTERNAL WORKING EXAMPLE. Built to show the instrument, not to assess a hotel or a person.
No property, person, owner or lender is named. Findings are properties of files.
Property consent is required before any specific below is quoted client-facing.
Solution 3 of 4 · internal working example · not for circulation
The Corporate Challenge Log
Every question corporate asks, typed and timed. The recurring ones become checks.
One log per cycle, kept corporate-side, recording every question corporate puts back on a pack: what was asked, what type of question it really is, how long it took to close, and how it was resolved. After two cycles the questions that recur become checks the control sheet prints, and content the pack answers before it is asked.
Built from: interview testimony, four sites Status: specimen, not a worked example Needs from Aleph: one seat at one review Page: v01 · 26 July 2026
Prepared by DH analyst · date
Reviewed by second person, never the preparer · date
Released by Aleph corporate · date
Specimen, not a worked example. DH holds no corporate correspondence: no email threads, no numbered challenge lists, no question logs. So the rows below are reconstructed from what four site interviews described, and are labelled as testimony throughout. This is the one solution of the four that cannot run until Aleph supplies something, and that something is small: a seat at one corporate review.
Evidence basis
4
Distinct corporate loop patterns described across four sites
0
Corporate questions DH has ever seen in writing
1
Thing needed to start: a seat at one review
2
Cycles before recurring questions become checks
The correction loop is the only inspection standing between a property and its owner, and it is entirely unrecorded. Nothing about it is written down anywhere in the estate, so the same questions are answered from scratch every month at every property, and nobody can say which question was asked twice. This page shows the instrument in shape and populates it with the loop patterns four sites described. Every row is testimony, not observation. Where a figure or a turnaround time was mentioned in an interview it is not repeated here, because in at least one case the number came from the interviewer rather than the witness, and an unanchored number on a page like this would be exactly the error the programme exists to remove.
Internal working example · properties anonymised · not for circulation
The log, in shape
Specimen rows, reconstructed from interview testimony at four sites and anonymised. The columns are the deliverable; the rows are illustrative of the patterns described.
ID
The question, as described
Type
What it tells the future state
Basis
CQ-01
A numbered list of challenges returned by email after submission, commercial in nature rather than corrective: cost lines and flow-through, not arithmetic.
Driver explained
Corporate is not checking the sums. It is asking why the business did what it did. A checker that only verifies arithmetic answers a question nobody is asking.
Testimony
CQ-02
Questions asked on the profit and loss statement rather than on the pack, answered by returning to source documents, principally the trial balance.
Number checked
The pack is not where the conversation happens. The ledger is. This is the strongest argument in the programme for obtaining a trial balance.
Testimony
CQ-03
At a cluster, corrections emailed to a finance team that sits outside the reviewer's reporting line, with drafted tables returned for correction before commentary can be written.
Document produced
The loop is not always property to corporate. Where a regional tier exists it runs inside the region, and the reviewer has no authority over the people he is correcting.
Testimony
CQ-04
At one site, no corporate loop was reported at all: where figures broadly align, the report is accepted as it stands.
None raised
Silence is not the same as quality. A log records the absence of questions as well as their presence, which is the only way to tell a clean pack from an unread one.
Testimony
The question taxonomy
The durable asset. Classifying by what a question is really asking for is what lets recurring questions become checks, and eventually content the pack supplies before anyone asks.
Type
What is really being asked
Where it goes after two cycles
Number checked
Is this figure right, and does it tie to the ledger.
Becomes an AR or SD check in the control sheet register, run before issue.
Driver explained
Why did the business do this, not what did it do.
Becomes required commentary: the pack answers it in the month it happens.
Decision justified
Why did you choose that, and what will you do now.
Becomes a commitment on the commitments page, carried until closed.
Document produced
Send me the thing behind the number.
Becomes a standing attachment, or an argument for a direct feed.
None raised
Nothing came back this cycle.
Recorded as a reading in its own right, so a quiet month is evidence rather than an absence of it.
How it is produced
Input: attendance at one corporate review, plus any question that arrives by email thereafter.
Method: log what is asked as it is asked. Classify by type. Record the date raised and the date closed.
Effort: two days to set up and sit the first review, then about thirty minutes per property per cycle.
The one ask of Aleph: a calendar invite to the August review. Request it from the sponsor rather than from the finance team, and log what is observed rather than waiting for anything to be forwarded.
Cadence: every cycle. The payoff arrives at cycle three, when the recurring questions have been identified.
What this does not do
This is a specimen. No row is observed. Every row is reconstructed from what a site described in an interview, and is labelled as such.
It cannot start without Aleph. It is the only one of the four solutions that needs something from the client: attendance at one review.
It does not measure turnaround yet. No response time or round-trip count has been obtained at any site, and the one figure mentioned in an interview came from the interviewer rather than the witness, so it is not repeated here.
It is a proxy, not the real thing. The charter's test is the owner's questions. Until owners are spoken to, corporate's questions are the closest available stand-in and should be described that way.
It names no one. Questions are logged by type and by tier, never by the person who asked them.
What it proves
The charter's test of a good review is that the owner reads it and does not need a call. Nobody can design toward that test without knowing what gets asked. This log is the cheapest available route to that knowledge, it stops the control sheet's check list being DH's taste rather than Aleph's, and it is the only one of the four solutions that compounds into pack content: a question asked twice becomes a question the pack answers before it is asked.
Notes and abbreviations
Every code used above, expanded. This block appears on every solution page.
Basis values used above
CQ-01 to CQ-04.
Challenge question, numbered. The permanent id of an entry in this log, so a question asked twice can be recognised as the same question and promoted into a check.
Testimony.
Described by someone in a recorded interview. Not observed by DH, not confirmed against a document, and labelled as such wherever it appears.
Observed.
Watched happen, or read out of a document DH holds. No row on this page carries this basis, which is why the page is a specimen.
Specimen.
The instrument shown in shape with illustrative content, as distinct from a worked example, which is populated from real files.
Question types
Number checked.
A request to verify a figure.
Driver explained.
A request to explain why something happened, rather than to restate what happened.
Decision justified.
A request to explain a choice and say what will be done next.
Document produced.
A request for the underlying document behind a figure.
None raised.
No question came back that cycle, recorded deliberately so that a quiet month is evidence rather than a gap in the record.
Cross-references
AR check.
An arithmetic and internal-consistency check in the control sheet's register. See solution 1.
SD check.
A check of the pack against Aleph's own published standards, in the control sheet's register. See solution 1.
Control sheet.
Solution 1: the one-page pre-issue check run corporate-side before a pack is released to its owner.
Commitments page.
Solution 2: what last month's pack promised and what this month's pack did with it.
Regional tier.
A layer between property and corporate that consolidates several hotels into one review. It exists in two regions of this estate.
Terms used above
BRM.
Business Review Meeting, the monthly pack and review an operator gives a hotel's owner.
Pack.
The monthly document a property produces for its owner. In this estate it is usually a PowerPoint presentation; at one cluster it is an Excel workbook.
P&L.
Profit and loss statement.
GOP.
Gross operating profit, before ownership costs such as rent, interest and depreciation.
ADR.
Average daily rate, the average room rate achieved.
Trial balance.
The listing of every ledger account and its balance. It is what a figure must be tied to before it can be called verified.
DH.
Deep Hospitality.
s22, s11.
Slide 22, slide 11, and so on, in the pack under review.