Toolbox Talk & Briefing Register

What this tool does

CM8-42 is a register of safety briefings and the people who attended them. It records what was delivered, when, by whom and to whom; it counts the talks properly rather than counting rows; it shows who has not been briefed recently; and it prints an attendance record you can hand to an auditor, a client or an investigator.

The reason it exists is uncomfortable but simple. After an incident, the first question is whether the person had been briefed on that hazard, and the second is whether you can prove it. Most organisations deliver the talks. Most cannot find the sheets. A talk that cannot be evidenced is, for practical purposes, a talk that did not happen.

Everything runs inside this single file. There is no account, no upload and no network request of any kind, so names, roles and attendance records never leave the computer you are using.

One row per attendee

This is the single most important thing to understand about the register, and it is where people go wrong.

The alternative — one row per talk with a list of names typed into a text box — is quicker to enter and useless afterwards. You cannot count how many talks a person has attended, you cannot find who was missed, and you cannot search for one name across a year of records. Attendance evidence is per person, so the register is per person.

Rows are grouped into talks by date plus topic. Two rows with the same date and topic are the same talk. The same topic on two dates is two talks — which is exactly what you want when a briefing is repeated for the night shift. It also means the topic must be typed identically each time: "Manual handling" and "Manual Handling" match, because case and extra spaces are ignored, but "Manual handling — loads" and "Manual handling" do not. Copy the topic rather than retyping it.

The tool refuses to record the same person twice on the same talk, and refuses two different durations for one talk, because both silently corrupt the figures.

What to record

Record every briefing where you would want the evidence later — the five-minute talk before an unusual job as well as the scheduled monthly session. In particular:

• Toolbox talks — short, task-focused sessions delivered at or near the work.
• Shift handovers that carried a safety point, where you would want to prove it was passed on.
• Briefings on change — a new machine, a new substance, a revised procedure, a new layout.
• Incident learning delivered after something went wrong, including near misses.
• Contractor and visitor inductions, if they cover the same hazards as your own staff.

Record the people who were actually briefed, not the crew list. If two were away and were briefed separately three days later, that is a second talk on a second date, not two extra names against the first. Entering people who were not there is the one thing that turns this register from evidence into a liability.

Categories

The category classifies the talk, not the attendee, so every row of one talk should carry the same one. The donut counts distinct talks per category — a fair picture of what your programme is actually about.

• Category — Use it for
• General safety — Recurring hazards everybody faces — manual handling, housekeeping, slips
• Task specific — A particular job about to be done, briefed at the point of work
• Incident learning — Delivered because something went wrong, here or elsewhere
• Seasonal — Weather, daylight, temperature, workload peaks
• New equipment — A tool, machine or substance being introduced
• Regulatory — A rule, policy or requirement that has changed
• Environmental — Waste, spills, emissions, storage
• Quality — Defects, checks and right-first-time work

A register that is nine-tenths general safety usually means the programme is running on a standing list rather than on what is happening in the business.

Counting talks, not rows

The headline figures deliberately separate the two things people confuse. Counting rows and calling it "talks delivered" overstates your programme by roughly the size of your crew.

Names are matched after trimming spaces and ignoring case, so "R. Osei" and "r. osei " are one person. They are not matched on anything cleverer than that: "R. Osei" and "Rita Osei" are two people as far as this tool is concerned. Pick one form of each name and stay with it.

The average is a plain arithmetic mean over the talks in the current filter. It is worth reading next to the minimum attendance target you set on the Settings tab: the talk summary marks any talk that fell below it, and a talk well below your normal crew size almost always means people were missed rather than that fewer were working.

Unsigned attendances are counted as a headline figure because they are the ones that fail an audit. A row with no signature is still a record that the person was briefed — it is simply a record with nothing behind it.

Briefing hours

Duration is the length of the talk, not the time per person, so it must be identical on every row of one talk. The tool enforces that. Two different figures on the same talk would either double-count the delivery or understate the person-hours, depending on which way you totalled them.

The tile shows delivery hours, with person-hours underneath, because the two answer different questions. Delivery hours tell you how much presenter time the programme consumed. Person-hours tell you what it cost in production time — nine talks of twenty-odd minutes each is under four hours of delivery but well over twenty hours of paid attention, and that is the figure to quote when somebody asks whether briefings are worth it.

If the same talk has been recorded with different durations because a row was edited, the longest figure is used and the register will have warned you at entry. Rows with no duration contribute nothing to either total, so the hours understate the programme rather than guessing.

Signatures and competency

Two separate ticks, and they are not the same thing.

Signed records that the person put their name on the attendance sheet. It is a record that a signature exists on paper or in your document system — it is not a signature, and ticking it does not create one. Keep the sheet. This register tells you which sheets should exist and which are missing; it is a index to your evidence, not the evidence itself.

Competency confirmed records that the presenter satisfied themselves the person can now do the thing safely — not merely that they sat through it. For most general safety talks it will be untouched, and that is correct. For a new tool or a changed procedure it matters: attendance means someone was in the room, competency means someone watched them do it. If you tick competency on every row by reflex, the field stops carrying any information at all.

Coverage and the frequency target

Set a briefing frequency target on the Settings tab — the longest anybody should go between briefings. The coverage table works out, for each person in the register, how long it has been since their most recent talk and flags anyone past it.

The coverage chart draws people fewest-talks-first, with anyone past the target in red, because the useful question is who is at the bottom of the list rather than who is at the top. Somebody who has attended one talk in six months is either being missed systematically — a night shift, a mobile crew, a part-time worker — or is simply not being included in the invitations. Both are fixable once you can see it.

Note that the target is compared against the calendar, not against the talks that were relevant to that person. Being briefed thirty days ago on a topic that has nothing to do with your job counts as current here. Use the number as a prompt, not a verdict.

What this tool cannot tell you

Be clear about the boundary, because this is a record with evidential weight and overstating it is worse than having no record at all.

• It cannot tell you who was absent. The register only knows the names typed into it. Coverage is calculated over people who have attended at least once, so anybody who has never been briefed does not appear anywhere in this tool — not in the coverage table, not in the chart, not in the totals. Read the coverage list against your own roster and look for the names that are missing from it.
• It cannot tell you whether the talk was any good, whether it was understood, or whether it was delivered at all. It records that somebody said it was.
• It cannot tell you what you are required to brief on, how often, or for how long the records must be kept. That differs by country, industry and contract.
• It is not a signature system. The signed tick is an index entry pointing at a sheet you keep somewhere else.

Printing and sharing

Print Report produces a report from whatever the current filter shows: the header you set on the Settings tab, the headline figures, the four charts, the talk summary, the coverage table, the full attendance register and your closing notes. Print to PDF to circulate or archive it.

The scope line under the title states the filter in force, so a report covering one quarter says so. Clear the filters before issuing anything described as the full record, and check that the date range matches the period you are claiming to cover — a filtered register with an unfiltered title is the easiest way to mislead somebody by accident.

Because the register names individuals and records whether they were confirmed competent, treat exports as confidential personnel information rather than as general site paperwork.

Saving your work

Records, settings and the report header are written to this browser's local storage as you type, and the toolbar shows the time of the last save. That storage belongs to one browser on one computer: another browser, a private window, a second machine or a clean-up tool that clears site data will not have it.

Treat Export .json as the real save — one file containing everything, which Import .json restores anywhere. Export CSV gives you the register for spreadsheet work and includes every filtered row, not only those drawn on screen, which is the export to keep when attendance records have to be retained for years. Reset asks twice, then erases everything this tool has stored. There is no undo.

Accuracy & disclaimer

This tool records what you enter and calculates from it. It cannot tell whether a talk was delivered, whether the people listed were present, whether the signature exists, or whether the competency tick was earned. Every figure on every screen depends entirely on the honesty and completeness of the entries.

Requirements for briefing content, frequency, records and retention differ by country and by industry, as do the rules on holding personnel information. This is an internal record-keeping aid, not legal advice, not a training management system, and not a substitute for the signed attendance sheets themselves.