Checklist
A practical guide to choosing a small representative batch, preparing source material, and reviewing output quality.
Checklist · 22 min
A practical guide to choosing a small representative batch, preparing source material, and reviewing output quality.
A useful evaluation tests the whole workflow from source material to approval.
Boardmate guide
The checklist shows which meetings to include, what source material to prepare, and how Boardmate drafts can be reviewed against the current workflow.
Share it with board support, the chair, and anyone signing off confidential sample material.
Guide sections
How to test Boardmate
Use this when a board support lead, chair, or service provider team wants to test Boardmate Lite with real working material before moving a wider backlog or live board cycle into the product.
Result of the exercise
Sample list, source assumptions, preflight result, draft outputs, reviewer activity, and open questions.
Proceed, retest, or pause, with the reason tied to quality, control, record, security, or timing.
A wider Lite backlog, a Full board cycle discussion, or a narrower retest with better source material.
Boardmate view
A rushed evaluation usually starts with one clean meeting, a quick upload, and a polite conversation about whether the draft looks good. That does not tell a chair, company secretary, fund administrator, or service provider team enough. Minutes are judged through source quality, reviewer control, chair decisions, output usefulness, and the record left behind.
The first Boardmate sample can be small but deliberately representative. It can include the ordinary work that happens every month, one or two cases that expose judgement, and at least one imperfect meeting where source material or reviewer behaviour is messy. The team sees the approval work, not only the first draft.
Evaluation detail
The weakest evaluation is also the easiest one to set up: send one tidy recording, admire or criticise the first draft, and then try to infer whether the product would survive real board support work. That produces opinions, but it rarely produces a decision a chair, administrator, legal reviewer, or procurement lead can rely on.
A stronger first sample has a deliberate shape. It tests normal repetition, one hard judgement case, one awkward source problem, and the review behaviour that usually slows minutes down. Boardmate is then judged as a controlled workflow from raw meeting material through approval, with draft speed only one part of the record.
The team knows whether the sample is testing backlog clearance, house-style quality, reviewer discipline, security review, or the case for a wider Boardmate Full conversation.
Include a messy meeting, a short meeting, and a complex meeting. If the product only sees the cleanest file, the team learns little about the real workload.
The evaluation ends with sample scope, source assumptions, review activity, output checks, unresolved questions, and a named next decision.
How to run it
Each pass creates a review record the next colleague can inspect. That is what makes the evaluation more useful than a product tour.
Before any files move, record the team problem, the sample scope, the review roles, and the decision maker. The brief names what counts as success, retest, pause, or wider rollout.
Use ordinary meetings to show repeatability, a complex meeting to test judgement, and a messy meeting to expose source handling. Prior approved minutes travel with the sample where possible.
A good product shows which rows and files are ready, missing, ambiguous, or recoverable. That is part of the workflow value, especially for backlog work.
Ask reviewers for selected-passage comments, uploads where needed, or no-comment confirmation. The chair or delegate decides what becomes source material for regeneration.
Open the DOCX, check the PDF, review any actions, and read the audit history. The team can explain how the final version was reached without searching inboxes.
Worked scenario
A Cayman fund board team chooses seven meetings: three routine quarterlies, one investment decision meeting, one meeting with sensitive director discussion, one adjourned meeting, and one meeting with late administrator notes. They include two current approved minutes and one older minute that is marked as historical style.
The team does not ask Boardmate to answer everything. It asks for a clear answer to four questions: can the first draft be reviewed as real work, can reviewers use the product without email sprawl, can the chair control regeneration, and does the exported file reduce final handling.
The sample shows repeated language, judgement, incomplete source material, and reviewer behaviour.
The sample shows one tidy meeting with no real review round and no DOCX inspection.
Board support prepares the pack; the chair or delegate controls accepted feedback.
Wider Lite batch, narrower retest, security pause, or Full Boardmate discussion.
A good result
The draft reads like formal minutes that can enter real review.
Reviewer feedback is visible, attributable, moderated, and consumed deliberately.
DOCX, PDF, actions, and audit history can support the team's evaluation file.
The team can explain the decision to a chair, IT reviewer, legal reviewer, or sponsor.
Boardmate review
Five to ten meetings gives enough variation to test ordinary minutes, complex judgement, short or adjourned meetings, and imperfect source material.
The CSV, source folder, previous minutes, board packs, transcripts, support notes, and exclusions are assessed before generation.
Reviewer comments, uploads, no-comment confirmations, chair instructions, and structure decisions remain visible before regeneration.
The evaluation is judged on DOCX usefulness, PDF reading copy, actions where available, and an audit history that explains how the file was reached.
Worked example
A useful first sample might contain seven meetings: three ordinary quarterly meetings, one administrator-report-heavy meeting, one meeting with sensitive director discussion, one short adjourned meeting, and one meeting with late support notes. The point is not to impress Boardmate with the cleanest material. The point is to expose the work a board support team actually has to do.
Prior approved minutes for the same entities help Boardmate match the expected record. Those minutes show attendance treatment, apologies, resolutions, recurring boilerplate, matters arising, defined terms, and the level of detail the chair normally accepts. If the prior examples are old style or from a different administrator, that limitation should be named before the draft is judged.
Confidentiality, sample scope, source assumptions, reviewer list, and security review questions are agreed.
Draft quality is read beside preflight notes, source assumptions, reviewer comments, and chair decisions.
The team keeps a decision note: proceed, retest, pause, or widen, with the reason recorded.
Decision ledger
Does this sample represent our real board support work?
Meeting mix, source list, prior-minute style pack, and explicit exclusions.
Proceed only if the sample tests routine work and at least one hard case.
Can we see source problems before they become drafting problems?
Rows, duplicate warnings, attendance checks, source-file matching, and missing-material notes.
Fix blockers, accept documented limitations, or narrow the sample.
Is the first draft close enough for real review?
Formal structure, prior-minute match, decisions, resolutions, actions, and restrained record language.
Review as real work, retest with better style reference, or pause quality assessment.
Can feedback become controlled input rather than inbox noise?
Private links, selected-passage comments, uploads, no-comment confirmations, accepted feedback, and chair instructions.
Regenerate only from accepted inputs and record what was rejected or held.
Can another colleague understand the result later?
DOCX, PDF, actions, audit record, open questions, and sample decision note.
Move to wider Lite, discuss Full, retest, or stop.
Watch points
Clean-demo bias One tidy meeting can flatter any product and still say little about backlog reality.
Authority drift If nobody decides which comments shape regeneration, every reviewer becomes an accidental editor.
Late security review Uploading confidential material before scope, access, retention, and deletion are agreed weakens the evaluation.
Boardroom wording
To the chair We are testing whether you can decide the next draft from a visible feedback set.
To board support We are measuring source preparation, review effort, DOCX repair, and final-record confidence.
To IT or legal No wider archive moves until sample scope, access, data flow, retention, and deletion are understood.
Boardmate toolkit
This gives the team a practical plan for a first Boardmate evaluation: what to prepare, what to observe, what to decide, and what to share with colleagues joining the decision later.
Briefing note
Before files move, the lead names the sample, the working question, the source assumptions, the review process, the chair decision point, and the close-out standard. That short brief prevents the sample from becoming a general reaction to one draft.
Backlog clearance, house-style quality, reviewer control, security readiness, or Full-platform suitability.
Meetings, entities, source types, reviewers, exclusions, exports, retention, and open security review questions.
Proceed, retest, pause, or widen, with one named person for the next action.
The meeting mix and file assumptions give preflight something defensible to check. Five to ten meetings is useful only when the sample explains why each meeting is there.
Read preflight before prose quality. Source ambiguity discovered early is product record, not an inconvenience.
Reviewer comments, support uploads, no-comment confirmations, held points, and chair decisions show whether feedback can be controlled.
The final review opens the working file, checks the reading copy, inspects action extraction where available, and reads the audit history beside the output.
Role by role
Prepare the sample matrix and source folder, then judge whether the output reduces the work around final minutes.
Sample list, preflight notes, DOCX handling, reviewer activity, and time saved.Decide which feedback changes the record before regeneration.
Accepted comments, held comments, rejected comments, and chair instructions.Review data flow, access, retention, deletion, and confidentiality before real material scales.
Scope note, user list, reviewer links, export record, and close-out decision.Use the review note to choose the next commercial step.
Proceed, retest, pause, or widen, with the reason tied to the sample.Decision route
One tidy meeting, no style pack, no real review round, and no export inspection.
Do not infer wider suitability from the sample.Several meetings, visible source assumptions, a formal first draft, and at least one reviewer experience tested.
Use the result for a narrow retest or a wider Lite discussion.Representative sample, chair-moderated feedback, useful DOCX output, and audit history that explains the final draft.
Move toward a wider backlog tranche or security review plan.Security questions, role responsibility, source naming, review policy, and close-out decisions are reusable.
Treat the sample as the working pattern for the next batch.Record note
This is the decision note the team can write after the first evaluation. It is deliberately practical, because the next colleague needs a record rather than a memory of a demo.
Seven meetings across routine, complex, short, adjourned, and imperfect source cases.
Preflight exposed source issues, generated reviewable drafts, kept feedback attributable, and exported usable files.
Security appendix, deletion process, one house-style retest, and confirmation of the chair decision maker.
Run a wider Lite tranche, or pause until the named blocker is resolved.
Avoid
Polished demo A polished walk-through with one clean file does not show backlog readiness.
Style memory Reviewers judging house style from recollection will disagree without approved examples.
Chair decision missing If accepted feedback has no decision maker, regeneration becomes another uncontrolled draft.
Late procurement Security questions handled after upload make the sample feel less controlled than the current process.
Set the bar
The evaluation should start with the working decision the team is trying to make. A backlog team may need to know whether scattered source files can become reviewed DOCX minutes. A fund administrator may care about repeatability across entities. A chair may care about control over accepted comments. IT or legal may care about access, retention, and audit trail before confidential material is shared.
Write those decision criteria down before the first upload. If the team only says that the draft should be good, the review will become subjective very quickly. A better standard asks whether the draft is reviewable, whether the reviewer experience is credible, whether the chair can moderate feedback, whether the exported files are practical, and whether the final record explains itself.
Which meetings were included and why they represent normal work.
What source material was available, missing, excluded, or added after preflight.
Proceed, retest, pause, or widen, with the reason recorded.
Choose material
The most useful first sample usually contains five to ten meetings. That is enough to show repetition, style patterns, exceptions, and review load without asking the team to examine an archive. Include ordinary meetings because they show routine value. Include a complex meeting because it tests judgement. Include a short or adjourned meeting because the product should not invent substance. Include an imperfect meeting because backlog work often starts with incomplete source material.
Prior approved minutes matter. They show how the organisation records attendance, declarations, approvals, matters arising, resolutions, deferrals, sensitive discussion, and boilerplate. If there has been a recent style change, mark which examples are current. If prior minutes are unavailable, record that limitation rather than judging style against memory.
Routine monthly or quarterly board support work.
Papers, resolutions, reports, sensitivities, or multiple reviewers.
Short, adjourned, incomplete, late-supported, or hard to match.
Judge output
The first draft still matters. It should read like formal minutes rather than a transcript digest, and it should be close enough for a real reviewer to work with. Boardmate also shows what happens around that draft: source assumptions, selected-passage comments, support uploads, no-comment confirmations, chair decisions, and the material used for the next version.
The DOCX needs practical scrutiny. Open it, inspect the headings, numbering, spacing, tables, names, dates, and any formatting that a board support team would otherwise repair by hand. The PDF should be a stable reading copy. Action extraction, if present, should be checked against the source record. The audit trail should answer what happened without forcing someone back into inbox search.
How long drafting, chasing, reconciliation, and final filing take today.
Where source control, review handling, regeneration, and exports change the work.
What still needs style, security, access, or procurement record.
Close properly
A useful evaluation produces more than a good impression from the call. It leaves a short decision note that names the sample, the source assumptions, the draft quality, the reviewer behaviour, the output package, and the unresolved questions. This note gives colleagues something concrete to discuss with the chair, legal, IT, procurement, or a wider board support group.
If the sample is promising but imperfect, retest with a better source folder or a different meeting type. If security or retention questions are not ready, pause before uploading a wider archive. If the workflow is strong, use the results to decide whether the next step is a larger Lite backlog, a Full Boardmate discussion, or a narrow pilot with formal security review.
Questions to settle
Carry forward
Sample matrix, source assumptions, draft quality notes, review activity, exports, and audit history.
Proceed to wider Lite batch, retest a narrower sample, or move security review questions forward.
Board support or the chair's delegate, with IT and legal brought in before confidential scale-up.
Evaluation question
A Boardmate evaluation asks whether a controlled minutes workflow can save time while preserving the organisation's house style, reviewer discipline, and final-record confidence.
Does the draft read like formal minutes the team could review as real work?
Can reviewer feedback be accepted, rejected, or held before it changes the record?
Can the team explain which source material, comments, exports, and decisions led to the final version?
Sample design
The first sample should be small enough to review properly and varied enough to be meaningful. Five to ten meetings is usually a better test than one impressive demo file because it reveals repeated language, edge cases, missing source material, and reviewer behaviour.
Representative, reviewable, and narrow enough for a real side-by-side assessment.
Only the cleanest meeting, only one document type, or material nobody has time to review.
A broad confidential archive before security, retention, and scope have been agreed.
Source material
Boardmate works best when the batch has a clear manifest and the source folder can be checked by a human. The goal is a credible source set: enough context to draft, enough naming discipline to match files, and enough prior wording to test style.
Preflight
The first useful moment in an evaluation is the preflight review. It should show whether the manifest and folder are coherent before anyone judges the generated minutes. If matching is ambiguous, that is useful to discover early because the current manual workflow probably carries the same ambiguity invisibly.
The meeting row and source material clearly belong together.
The expected file is absent, incomplete, or intentionally out of scope.
A file may belong to more than one meeting, entity, or date and needs human confirmation.
The team can add a pack, transcript, prior minute, or note before draft generation.
Review standard
The output should be judged as a working governance record. A useful review checks the draft itself, reviewer access, chair decisions, and the export package together.
Decision
A strong evaluation leaves the team with a grounded view of where Boardmate changes the work: faster first drafts, fewer email reconciliations, clearer reviewer history, and a more complete final record. The decision should also name any remaining gaps before wider rollout.
Quality is close enough for review, time saved is material, and the governance workflow is clearer than the current process.
The workflow is promising but needs another sample, clearer source material, or stronger house-style examples.
Security, retention, access, or approval ownership is not yet settled enough for real board material.
Keep the sample meetings, source assumptions, reviewer activity, exports, and open questions together for team sign-off.
Common questions
Five to ten meetings is usually enough if the sample includes ordinary, complex, short, adjourned, and imperfect examples.
No. Missing and imperfect material is part of the test. The important step is to mark gaps clearly so the team can separate source issues from drafting quality.
The person who normally owns the minutes outcome should be involved, usually board support with the chair or delegated approver.