American workers quit their jobs roughly four and a half million times in February, according to figures published last week, against some eleven million open positions. The pattern has held for a year now, inflation is cutting real pay for anyone who stays put, and the easiest way to get a substantial rise in 2022 is to resign. Finance operations is feeling this differently from other functions, and the difference is worth naming precisely. When a marketing manager leaves, work slows. When an accounts payable supervisor or a group accountant leaves, a dated obligation fails in public: the close slips, a filing is late, a payment run does not go out, an auditor's request goes unanswered. Finance is the one back-office function with a hard periodic deadline, statutory penalties attached to it, and a control framework that specifically prohibits letting one person cover for another. That combination is why finance attrition costs more than its headcount suggests, and why the remedy is not primarily a recruitment remedy.
The close calendar is your real organisation chart
Here is a diagnostic that takes an afternoon and tells you more than any engagement survey. Take your month-end close calendar and your statutory filing calendar. For every task, write the name of the person who actually performs it — not the role, the person. Then write the name of the second person who could perform it unsupervised, today, without asking anybody. Count the tasks where the second column is empty. In a typical mid-sized group, between a quarter and a half of close tasks have nobody in the second column, and the empty ones cluster: intercompany elimination, the tax return, the payroll journal, the inventory provision, the bank reconciliation for the account nobody else has access to, the consolidation spreadsheet with the linked workbooks. That count is your exposure, expressed in the only unit that matters — tasks that will not happen if one person gives notice.
Write down the exceptions, not the process
Most knowledge-capture initiatives produce a process manual, and process manuals are not where the risk lives. The process is the generic part; it is in the system documentation and in the head of any competent accountant you hire. What leaves the building is the exception set. The customer whose invoices must be split by site because their portal rejects consolidated billing. The supplier who is paid in advance because of a dispute settled three years ago. The bank that rejects payment files submitted after early afternoon. The account that always shows a difference of a fixed amount because of a legacy migration entry nobody wants to write off. The intercompany balance that is reconciled manually because the two entities use different rounding. The government portal that requires a specific browser. Four artefacts capture almost all of it, and each one is small enough to maintain: A recurring-entry register — every standing journal, its purpose, its basis of calculation, its approver, and the date it should stop. A reconciliation pack with a known-differences list — for each account, the expected reconciling items and why they exist. An exceptions log — counterparty by counterparty, every deviation from the standard process and its origin. An access and credentials map — which portals, tokens, certificates and signatory authorities each person holds, for which entity, and who the backup is. None of that is a documentation project. All of it is a byproduct of doing the work, if someone insists on it being written where the next person will look.
| Artefact | What it preserves |
|---|---|
| Recurring-entry register | Purpose, basis, approver and stopping date. |
| Reconciliation pack | Known differences and their explanations. |
| Exceptions log | Counterparty deviations and their origin. |
| Access map | Portal and authority owners, approved recovery and backups. |
Qualitative summary of this article's source text, not a measured outcome or performance estimate.
Record the close while it happens
The cheapest form of capture is a screen recording made during the actual close, with narration, task by task. Twenty minutes of somebody doing the consolidation and saying what they are doing is worth more than forty pages of prose, costs nothing to produce, and is honest about what the work involves — including the copy-paste steps the official procedure does not mention. Index the recordings against the close calendar so they are findable by task rather than by date, and accept that they will go stale. A slightly stale recording of how the job was really done in March beats an immaculate manual describing how it was supposed to be done in 2019.
Assume the notice period will be shorter than you think
People leave badly, get bought out of notice, or become unavailable in the last fortnight. Build a handover template that front-loads the irreversible items into the first week: system access inventory, credentials held personally, in-flight transactions and queries, the diary of dated obligations for the next ninety days, the list of people at customers, suppliers, banks and authorities who need to know, and anything sitting in a personal mailbox that is the only record of an agreement. The last item is the one that hurts most. In finance operations an enormous amount of consequential agreement — a credit note approved, a payment deferred, a variance explained — exists as an email thread in one person's mailbox. A departure makes it unreachable at exactly the moment an auditor asks.
What actually retains finance people
Money first, honestly. Inflation has repriced the market and internal increment cycles have not caught up, so an off-cycle correction for named critical roles is cheaper than a vacancy. But two non-monetary levers work unusually well in finance because the pain is unusually concentrated. The first is close-week relief: shortening the close, moving tasks earlier, automating the reconciliations that cause the twelve-hour days. The grievance in most finance teams is not the average week, it is the same five days every month. The second is qualification support with actual study leave, which buys loyalty through the exam cycle and signals a path. What does not work is a retention bonus paid after the resignation. By then the decision is emotional as well as financial, and you have taught the team the mechanism for getting a raise.
Practical Guidance for Process Knowledge Capture
- Map every close and filing task to a first and second name, and publish the count of gaps to leadership.
- Capture exceptions, credentials and recurring entries before attempting any process documentation.
- Record the close as it happens, indexed by task, and keep the recordings where the team already looks.
- Move consequential email out of personal mailboxes into shared records for anything approving a financial treatment.
- Rotate one high-risk task per month so second names become real rather than nominal.
- Front-load handovers into the first week of notice, assuming the rest will be lost.
- Correct pay compression off-cycle for critical roles, rather than waiting for the annual review.
- Attack the close-week workload, because predictability retains finance staff better than a token increase.
The Regional Angle
Three regional developments change this calculation in 2022, and the first arrived two months ago. The new Emirati employment law came into force at the beginning of February, and it makes movement easier in ways that matter for retention. Fixed-term contracts became the standard form, notice and probation rules were tightened and clarified, and — significantly — an employee in probation can move to another employer in-country on short notice. Alongside it sits a new category of permit for part-time, temporary and freelance work. The received wisdom that regional attrition is structurally low because residency is tied to a single employer is weakening in real time. It also hands you an instrument almost nobody is using yet: the group accountant relocating with a spouse, or the controller retiring, can legitimately be retained part-time for the close week under the new permits. Ask about that before accepting a resignation as total. The second is a skills shock that will arrive from the autumn. Federal corporate tax was announced at the end of January, applying to financial years starting in mid-2023, in a jurisdiction that has never had it. Every group will need people who can handle a corporate tax computation, transfer pricing documentation and a deferred tax position — on top of value added tax, economic substance filings and, for Saudi entities, the next phase of electronic invoicing. The consequence for retention is direct: the market is about to bid aggressively for exactly the mid-career accountants you are trying to keep, and your close calendar is about to acquire new tasks with nobody in the second column. Identify who in your team will own the corporate tax workstream now, fund their training this year, and treat it as a retention conversation rather than a project resource question. The third is about the bridge you will reach for when someone leaves. The default regional response is to ask the audit firm for a secondee from its outsourcing arm, because the relationship exists and mobilisation is fast. That is frequently the right answer commercially and a problem professionally: if the same firm audits the entity, having their people prepare the records they will later audit creates an independence issue your audit partner will eventually have to raise, usually at the least convenient moment. Line up an alternative bookkeeping and interim-resourcing provider before you need one, and keep the auditor for audit.
The objection worth taking seriously
The strongest objection comes in two parts. First, the Great Resignation is an American story that gets over-read everywhere else; in the Gulf, employment and residency have been linked closely enough that voluntary attrition has historically run well below Western levels, and extrapolating from monthly quits data published in Washington is lazy. Second, knowledge capture is where good intentions go to die. Documentation rots within two quarters, recordings are never watched, the exceptions log is updated for three months and abandoned, and the honest answer to losing good people is to pay the market rate and hire competent replacements who bring their own methods. Both parts have force. Documentation decay is close to a law of nature, and paying market is the correct first-order response. But the regional premise is precisely what is changing: a new labour law that eases movement, a corporate tax regime creating a scarce skill, and sustained hiring by neighbouring markets have together loosened the constraint that used to suppress attrition here. And the capture that survives is not a documentation project — it is the small set of artefacts a team uses every month because they are the fastest way to do the work. The recurring-entry register gets maintained because the reviewer asks for it. The exceptions log stays current because payables consult it daily. If an artefact has no weekly reader, it will rot, and the conclusion to draw is not that capture is futile but that you chose the wrong artefacts.
Common Questions
What is the single most useful thing to capture?
The access and credentials map, followed by the exceptions log. Both are short, both are immediately actionable, and both are what fail first when someone leaves unexpectedly.
Should we outsource the close instead?
Outsourcing transfers execution, not knowledge, and it works only if the exceptions are documented — which returns you to the same task. It is a reasonable answer to a capacity problem and a poor answer to a knowledge problem.
How do we make second names real?
Rotate one high-risk task per close and have the primary person review rather than perform it. Cover that is never exercised is not cover.
What should we expect over the next twelve months?
Expect quits to stay elevated through the summer and to cool as higher interest rates slow hiring later in the year, without returning to 2019 levels. Expect finance pay to reprice most sharply in the three-to-seven-year experience band, which is where your close depends. Expect corporate tax readiness to become a competing bid for those same people from the autumn. Expect part-time and fractional finance arrangements to become normal under the new permit categories. And expect the first close after any senior departure to be the one that slips — which is the argument for doing the mapping exercise this month rather than after the next resignation.
Process Knowledge Capture — we map your close calendar to named people, capture the exceptions and credentials that actually leave with them, and build cover you have tested rather than assumed.