Review & Cadence

How to review a risk register on Mac

RiskOS gathers everything that is due into one queue and hands you the risks one at a time, worst first. Here is how to work a pass from end to end.

A register ages quietly. Owners move on, a planned control starts operating, a supplier is acquired, and the ratings argued over so carefully in March stop describing the organisation you have now. What keeps a review happening is rarely discipline. It is speed: a pass that takes a morning gets run every quarter, and one that takes a week gets postponed until somebody asks to see it.

Note

Everything here happens on your Mac. There is no account, nothing is uploaded, and your register never leaves the machine.

Where review mode lives

Choose Review in the sidebar, in the Register group beneath Risks, Controls and Actions. It opens on a scope picker rather than a table, because the first decision in any review is which risks this pass is actually for.

Each scope carries a live count, so the size of the job is visible before you commit to it. Choose one and RiskOS hands you the risks a single one at a time, worst first, with a progress bar showing how far through the queue you are. The scoring inputs sit beside the risk's context — owner, last review, linked controls, open actions — so a decision never rests on memory.

Work a review pass, start to finish

  1. Open review mode

    Choose Review in the sidebar. The scope picker appears with a count beside each option, telling you how many risks are overdue, how many fall due soon, and how many sit above the appetite that applies to them.

  2. Choose the scope for this pass

    Pick one of Overdue for review, Due within 30 days, Above appetite or All active risks. Choose the narrowest scope that answers the question you came to answer: a queue of six risks gets worked properly, and a queue of sixty gets skimmed.

  3. Read the context before you decide

    The first risk arrives with everything you need beside the ratings: who owns it, when it was last reviewed, which controls are linked and what state they are in, and what actions remain open. Read that before touching a stepper. Most reviews are settled by the context rather than by the arithmetic.

  4. Confirm a risk that has not moved

    If nothing about the exposure has changed, press to confirm. The review date is stamped, the next review is scheduled from the risk's cadence, and the queue moves on. Confirming is a decision in its own right, not a way of avoiding one.

  5. Re-rate a risk that has changed

    Where something has genuinely moved, adjust the inherent likelihood and impact on the steppers. The projected rating updates live as you work, so you can see where the risk would land before anything is written. If a linked control has started operating since the last pass, the residual figure answers on its own.

  6. Save the new rating and move on

    Press to save and go to the next risk. The new assessment is recorded, the review date is stamped, and the next review is scheduled from the cadence on the risk. Your hands never leave the keyboard between one risk and the next.

  7. Skip anything you cannot settle today

    When a risk needs a conversation you cannot have this morning, press to skip it. Skipping leaves the risk completely untouched, so it keeps its review date and turns up again in the next pass. steps back to the risk before.

  8. Work the queue to the end

    Keep going until the progress bar fills. If you need to break off and check something elsewhere in the register, go ahead: the session survives leaving it, so you can return to Review and carry on rather than starting the queue again.

  9. Read the summary at the end

    When the queue empties, a summary shows what moved, risk by risk. Read it before you close the session. It is the clearest picture you will get of what the pass changed, and the natural moment to export a report while the reasoning is fresh.

Choosing a scope for the pass

The four scopes answer four different questions, and mixing them is the quickest way to turn a focused hour into an aimless one. Pick the question first.

The four review scopes and when each one earns its place
ScopeWhat it gathersWhen to run it
Overdue for reviewRisks whose next review date has already passed.Monthly. This is the queue that should be empty and rarely is.
Due within 30 daysRisks falling due over the coming month.Ahead of a committee date, to clear the month before it clears you.
Above appetiteRisks whose residual score sits above the threshold that applies to them.Before a board pack, so nothing over the line arrives unexamined.
All active risksEvery risk still open in the register.Annually, or after a change big enough to move several ratings at once.

Run the overdue pass first

An overdue risk is one where the register has stopped being evidence of anything. A rating last checked eight months ago tells a reader that nobody has looked, which is worse than an uncomfortable rating checked last week. Clear this scope before widening to any other, and often enough that it never grows past an hour's work.

Keep the appetite pass separate

The appetite scope is not maintenance. It is the conversation about what you are prepared to live with, and it belongs on its own, with the people who set the threshold in the room. Running it separately also stops the overdue queue being derailed by the two or three rows that need a decision rather than a date.

The keys that do the work

Review is built to be worked from the keyboard, because the gap between a review that gets finished and one that gets abandoned is usually the number of clicks per risk. Four keys cover the whole pass.

Review mode keyboard shortcuts and what each one records
KeyWhat it doesWhat it records
Confirm the risk as it standsStamps the review date and schedules the next one from the risk's cadence.
Save what you changed and move to the next riskRecords the new assessment, stamps the date, schedules the next review.
Skip this riskNothing at all. The risk is left exactly as it was.
Go back to the previous riskNothing. It is a way of re-reading, not of undoing.

Z undoes a change made by mistake, and Z puts it back.

What the register records when you decide

Three outcomes, three different marks left behind. Knowing which is which is what lets you trust a review date months later.

A confirmation is a decision

Confirming says the rating still holds. It stamps the review date and schedules the next review from the cadence set on the risk, so the row leaves the overdue queue and returns when it is genuinely due. A re-score that ends up changing nothing is treated the same way and recorded as a confirmation, which is the honest description of what happened: somebody looked, and the number stood.

A re-score writes history

A changed rating is kept as an assessment in the risk's History, with the score it produced and a chart of the residual over time. A year later the question is never "what is this rated?" but "when did it change, and what was going on when it did?" — and only the history answers that.

A skip writes nothing

Skipping is deliberately inert. No date is stamped, no assessment is written, and the risk keeps the review date it had, so it is still due the next time you open a queue. Use it freely: a skip costs nothing but this morning's place in the queue.

Reading the context beside the ratings

Everything sits on one screen because a rating changed without context is a guess with better presentation. Here is what to look at, and what each part is telling you.

The controls, and whether they are running

Linked controls appear with their status and effectiveness. Only controls that are implemented or operating reduce a risk, so a planned control that has sat unfinished through the last two reviews is holding nothing down, however strong it will be when it lands. If a control has started operating since you last looked, that alone may be the whole review.

The actions nobody has done

Open actions against the risk sit in the same view. Two overdue actions beside a rating that has not moved usually means the treatment has stalled, and the useful outcome of the review is a new owner or a new date rather than a new number.

The projected rating, before anything is written

As you move a stepper the projected rating updates: the score, its band, and where the risk would sit against its appetite threshold. Nothing is written until you save, so you can try a rating, see that it would carry the risk over appetite and have it flagged everywhere it appears, and decide whether that is what you mean.

Troubleshooting

The queue is empty, but I know risks are due

Check which scope you chose. Each carries its own live count, and if Overdue for review reads zero then nothing has passed its next review date. Switch to Due within 30 days to see what is coming, or All active risks to work the whole register.

A risk I skipped keeps coming back

That is the design. Skipping leaves the risk completely untouched, including its review date, so it stays due and rejoins the next pass. To take it out of the queue, decide something: confirm it, re-score it, or change its cadence in Risks.

The next review date is not the one I expected

The date is counted forward from the review you have recorded, using the cadence set on that risk rather than a register-wide default. Open the risk in Risks, look at Summary, and change the review cadence there. Review the risk again and the next date follows the new cadence.

I had to leave halfway through a session

Nothing is lost. The session survives leaving for another section, so you can read a control's evidence or check a vendor's review date and then return to Review and pick up where you stopped. Anything saved is already recorded, and anything skipped is still waiting.

The projected score is not what I expected

The residual figure comes from the effectiveness of the controls you have linked, unless a value has been set by hand on that risk. Check the control statuses first: a planned control reduces nothing yet. Where a hand-set effectiveness disagrees with the linked controls, RiskOS points that out in the risk's panel rather than quietly choosing one.

A review routine that holds

A review process survives a busy quarter only if it is short, repeatable and run the same way each time. These habits carry most of the weight.

  • Book the pass, not the outcome. An hour set aside every month for the overdue queue beats a day put by each year that keeps getting moved.
  • Set cadences you can meet. Cadence is what fills the queue. If every risk is monthly, the queue is always too long to work properly and the dates stop meaning anything.
  • Let worst first stay first. Risks arrive in order of severity, so if you run out of time you have spent it on the rows that matter most.
  • Fix the actions while you are there. An overdue action beside an unchanged rating is the review telling you where the treatment has stalled.
  • Export the record when the pass ends. A report made straight afterwards captures the register in the state the review left it, and RiskOS builds every format from that one snapshot, so they agree with each other.

Frequently asked questions

How often should a risk register be reviewed?

Set a cadence per risk rather than one for the whole register. Critical and fast-moving risks usually warrant monthly attention; steady operational risks are fine quarterly; low, stable risks can sit on an annual cycle. Each next review date is scheduled from the cadence on the risk, so the overdue queue stays a realistic morning's work rather than an annual reckoning.

How do I review only the risks that are overdue?

Choose Review in the sidebar and pick the Overdue for review scope. It carries a live count, so you can see how many risks have passed their next review date before you start. The queue then hands them over one at a time, worst first, with a progress bar showing how much of the pass is left.

What happens if I skip a risk during a review?

Nothing at all is written. Skipping with ⌘→ leaves the risk completely untouched, including its review date, so it is still due and turns up again in the next pass. That makes skipping safe: it defers a decision you cannot make this morning rather than quietly marking something as done.

Does confirming a risk without changing it still count as a review?

Yes. Confirming stamps the review date and schedules the next review from the risk's cadence, exactly as a re-score does. A re-score that ends up changing nothing is recorded as a confirmation too, because that is what happened: somebody looked at the risk and the rating stood. The evidence of a review is the date, not the movement.

Can I review a risk register using only the keyboard?

Yes, and it is the fastest way to work. ⇧⌘↩ confirms a risk, ↩ saves a re-score and moves to the next one, ⌘→ skips and ⌘← goes back. With those four you can work an entire queue without reaching for the mouse.

What is the difference between marking a risk reviewed and running a review pass?

Marking reviewed is a bulk action from the risks table: select several rows and stamp them at once, useful when a batch has genuinely been looked at together. A review pass takes one risk at a time with its controls, actions and history on screen, so the decision is made with the context in front of you rather than from a list.

Can I stop a review session and come back later?

Yes. The session survives leaving for another section, so you can check a control, look at a vendor's next review date or read an event, then return to Review and continue where you stopped. Everything saved or confirmed is already recorded, and anything skipped is still in the queue waiting.

Where can I see what a review actually changed?

Three places. The summary at the end of the session lists what moved, risk by risk. Each risk's History keeps every assessment with the score it produced and charts the residual over time. A report exported afterwards captures the whole register in the state the review left it, which is the version worth filing.