“Outsourced accounting is what you do when you cannot afford an in-house accountant.” I hear that constantly. It does specific damage, because it shapes how the owner engages with what they bought.

An owner who believes they have an interim, budget-constrained arrangement treats it that way. They do not invest time in the monthly call, because they do not think of the provider as someone worth investing time with. They do not ask about margin by client or cash flow by quarter, because they assume the function is not sophisticated enough to answer. They wait for the real thing while underusing the service they are paying for.

The structural advantage of the model is access, not price. A business at AED 5 million hiring an in-house accountant at AED 8,000 to AED 12,000 a month, the Abu Dhabi market rate for someone qualified, gets one person with one skill set. That person is unlikely to also be current on Corporate Tax law, IFRS for SMEs, sector-specific VAT treatment, voluntary disclosure mechanics and transfer pricing.

An outsourced firm at the same or lower cost gives access to a team: the bookkeeper, the reviewing accountant, the tax-aware senior, the manager overseeing the engagement. No single hire at that budget matches that breadth.

The genuine constraint is proximity. An in-house accountant is embedded in daily operations in a way an outsourced provider cannot fully replicate. That matters where the finance function needs to be in the room for commercial decisions continuously, which is closer to CFO services than to an operating function. For most mid-market UAE businesses, it is not what the finance function is for.

A management consulting firm, Abu Dhabi mainland, four fee-earners including the founder, revenue around AED 3.8 million. This is a composite drawn from a conversation I have regularly rather than a single named client, but the numbers are the ones that come up.

The owner was weighing an in-house accountant at AED 10,000 to AED 12,000 a month in salary, plus roughly AED 3,000 in visa, insurance and end-of-service accrual. Call it AED 13,000 to AED 15,000 a month all-in.

The recommendation was the third option. A four-fee-earner consulting firm does not need its finance function embedded in daily operations. The rare urgent situations, a bank query or FTA correspondence, are manageable through a responsive relationship with a defined escalation protocol.

The difference funded a significant part of a new fee-earner the business was also considering. That is the point worth holding onto: the comparison is rarely just about the finance function. It is about what the released budget does elsewhere.

Outsourced finance engagements arrive in two shapes, and the difference between them is timing rather than sector or scale.

The rescue is the common one, and it starts with backlog accounting. A property business came to me with no functioning finance function at all: no accountant, no bookkeeper, no accounting system in any meaningful sense. Three years of trading history existed as bank transactions, a folder of scanned invoices, and abandoned entries in a system nobody had maintained. The reconstruction ran transaction by transaction across three financial years before any ongoing function could begin.

What came out of it was the company’s first reliable set of accounts, ever. Three consecutive years of IFRS-compliant statements and, eventually, three consecutive statutory audits with clean opinions, the first in the company’s history.

The Foundation Model

The opposite profile arrived at exactly the right moment. A newly established engineering manpower consultancy contacted me before the disorder had a chance to develop. The owner had already made one good decision: adopting a cloud ERP from day one rather than starting with spreadsheets.

That was an advantage and a complication. An out-of-the-box chart of accounts is not configured for UAE requirements, and this sector has a specific structure. Revenue comes from placing technical staff, on time and materials or on fixed-term secondment, and the salaries of those staff are direct costs attributable to specific contracts rather than general overhead.

The setup used the platform’s analytic accounting to tag every transaction with a contract reference, producing a second-dimension profit and loss alongside the standard chart. Every invoice raised and every staff cost posted carried that reference from the first transaction.

The commercially important numbers there are utilisation, the percentage of placed staff on billable contracts, and margin per contract. Both are invisible without contract-level tagging, and both were visible from the first full month of trading.

The lesson I apply to every new engagement is that getting the setup right at the beginning costs a fraction of fixing it later. The chart of accounts that took a week to design at inception would have taken months to retrofit after two years on a default configuration. The VAT classification built into the invoicing workflow from day one would have needed a voluntary disclosure to correct if applied inconsistently across two years of returns.

The foundation is always cheaper than the rescue. The problem is that it looks like an overhead at the moment a business needs to spend money on everything else at once, while the rescue looks like a necessity once the pain is visible. That is why rescues are more common.

If the first three months are done properly, the engagement produces value from month four onward with relatively low ongoing effort. If they are done quickly or cheaply, it produces accurate records without insight, and embedded problems compound quietly until they surface at the worst moment.

The Access Audit

The first thing I ask is not which software you use. It is who currently has login access to EmaraTax, the bank portal, the accounting platform, the payroll system and the company email accounts.

Clients never think to ask this, because it feels administrative rather than strategic. It is the question that decides whether the transition is clean or chaotic.

What I find, repeatedly: the previous bookkeeper holds credentials the owner does not have. The EmaraTax account is registered to a personal email belonging to someone who no longer works there. The accounting platform has a single admin login under the previous bookkeeper’s own credentials, never transferred. The bank portal needs an OTP device in a former employee’s pocket.

Each is a specific operational risk at the moment of transition. If the new provider cannot access EmaraTax on day one, the VAT return due at month end cannot be filed. Every one of those resolves cleanly if identified before the engagement starts, and becomes a crisis if discovered three days before a deadline.

Verifying the Opening Balances, Not Just Entering Them

Every handover includes entering the opening balances. What most skip is testing whether those balances are correct.

The usual pattern: the previous arrangement produces a trial balance at the handover date, the new provider enters it and starts posting forward. Nobody tested whether bank balances agree to actual statements, whether debtors agree to a reconciled aged listing, or whether creditors agree to supplier statements. Nobody checked the VAT control account against VAT declared and paid across historical periods.

In most handovers I have conducted, the answer to at least one of those is no, and often to several at once.

The consequence is that errors compound rather than stay put. Take a debtor balance overstated by AED 80,000 at handover, because an invoice that should have been written off is still in receivables. That overstates profit by the same amount in the period the write-off is finally recognised, and the CT return built on those accounts carries a wrong taxable income figure.

My verification has four components, all completed before the first month closes. A bank reconciliation across every corporate account at the handover date, and a debtors verification against an aged listing, with anything past ninety days assessed for recoverability. A VAT control account reconciliation against cumulative VAT declared versus paid. And an end-of-service position calculated per employee and compared to the accounts.

End-of-service benefits are the most consistently mishandled accounting obligation I encounter, in every sector.

The requirement under IAS 19 is unambiguous. When an employee earns the right to a future benefit, the gratuity under UAE Labour Law, the employer recognises a liability then, not on payment. What I find almost universally is that it has been recorded only as a cash expense when someone leaves.

That understates the balance sheet liability throughout an employee’s tenure and overstates profit in the early years relative to the final one. For a business with ten employees averaging four years of service at a reasonable salary level, the unrecognised liability is typically AED 150,000 to AED 400,000.

Recognising it creates a charge to the income statement and a liability that changes the net asset position. The downstream effects reach the CT return, the balance sheet the bank sees, and audit scrutiny. All manageable when recognised proactively, and considerably less so when the auditor or the FTA finds it first.

The schedule I build into every engagement from the first month is one row per employee: name, start date, current basic salary, years of service, calculated liability. It updates on salary changes and produces the monthly provision charge. It is the first document an auditor asks for, and it takes about thirty minutes to set up.

The WPS Reconciliation Most Providers Skip

Where payroll is handled internally or by a separate processor while accounting sits elsewhere, the reconciliation between the two has to be explicitly managed. In most takeover engagements, I find it has never been done.

The gaps that emerge are rounding issues, one-off deductions applied in payroll but not posted in the accounts, and employer contributions omitted from the expense base. Differences can legitimately exist. They just need to be explained, because an unexplained difference between payroll expense and WPS transfer records is precisely what an audit treats as a flag. Fifteen minutes a month produces a clean payroll audit trail from the first cycle.

A fabrication and light manufacturing business in Mussafah, Abu Dhabi, roughly eight office staff and thirty workers on site, making metalwork and fit-out components for construction and infrastructure clients.

The owner’s requested scope covered bookkeeping, VAT filing, CT compliance, WPS payroll, annual leave tracking, invoice management, project costing and monthly reporting. Seven distinct components, which is a finance function rather than a bookkeeping service. His previous freelance bookkeeper had left him without project costing, without reliable monthly reporting, and without a payroll process that integrated correctly with WPS.

I should be transparent about the commercial side. My initial assessment put the right retainer at AED 4,500 to AED 5,500 given the scope, the workforce of thirty-eight and the job-costing complexity, while his budget was AED 3,000. I took the engagement at that level, with project costing implemented progressively rather than fully from day one. The alternative was that the most valuable component never got built.

The Component That Paid for Itself

Project costing was the structural decision that mattered. I set up job codes, one per active fabrication order, with every materials purchase coded to the relevant job and labour allocated through weekly site-supervisor timesheets. A week to design, fifteen minutes a month to maintain.

In the first month it ran, the owner identified two job categories priced consistently below their actual cost basis. Pricing was corrected the following month, and the margin improvement over the subsequent quarter was approximately AED 18,000. Six months of retainer recovered from a single pricing correction, on information that had been invisible for as long as the business had been trading.

The payroll component was the time-sensitive one. The MoHRE deadline is strict, and failure attracts a block on new work permit applications, which for a labour-dependent contractor is a material operational risk rather than an administrative one. The process set a cut-off for timesheet and absence data by the 25th, an approved payroll register by the 28th, and WPS submission by the 30th.

The most common and most costly mistake is treating the provider as a data processor. A processor receives information and records it accurately, which is valuable and is the minimum the engagement should produce. A finance function does that and also asks questions.

Why is this supplier being paid twice the amount on the invoice? Why has this receivable been outstanding for ninety days when the contract says thirty? Why did the margin drop four points with no obvious change in revenue? Why is there a payment to an entity nobody recognises?

The pattern I see is an owner who emails invoices and bank statements once a month, receives accounts two weeks later, glances at the bottom line, and considers it complete. The provider has done what was asked. Nobody asked the question.

The cost is specific. A margin drop nobody questioned for three months turned out to be a pricing error, a line item quoted at AED 35 that should have been AED 45. It was entirely recoverable from the client in month one, and considerably less so by the time it surfaced. Which happened because the owner compared two invoices in a meeting, not through any systematic review.

A ninety-day receivable nobody followed up became a write-off when the client went into liquidation. The amount was on the aged debtors report every single month.

The difference is not primarily the skill of the firm. It is the governance of the engagement.

If I could give one piece of advice to anyone managing this kind of relationship, it is to treat the monthly review call as the meeting you do not reschedule.

When it gets deprioritised, the accounts arrive by email, the owner checks the bottom line, decides it is roughly where expected, and closes the message. The provider has done its job. The conversation that should have happened around those accounts did not, and the conversation is where the value is.

A margin drop that should prompt a question about supplier costs or pricing goes unasked. An aged debtor at ninety-three days from a client the owner knows is cash-constrained goes undiscussed. A next-quarter VAT liability that is materially larger, because a prior-quarter input recovery is not repeating, goes unflagged. All of it is in the accounts, and none of it surfaces without the call.

The provider will not chase you for the meeting. It will send the accounts, follow up once, and move to the next client. That is how these relationships operate at scale, and it means whether the management information changes any decision depends entirely on whether the conversation happens.

What works is a standing date each month with a fixed agenda: management accounts, debtors and creditors, the VAT position, compliance matters and one commercial question. Thirty to forty-five minutes, with the person responsible for the engagement rather than a junior bookkeeper.

The selection criteria are different when a provider manages the platform rather than the owner. Multi-client management, reliable FAF export and UAE-specific configuration matter more than the feature list.

Zoho Books is the most common platform in my portfolio, largely because it is what clients arrive with and migration rarely justifies its cost. Configured correctly it produces reliable output, and its real limitation is multi-entity management. For a new UAE engagement starting from scratch, Wafeq is my current recommendation, on UAE-native configuration and Arabic support. Xero has the cleanest audit trail and the strongest integration ecosystem, which makes it right where the finance function connects to inventory or project management systems.

Disclosure. I am building a UAE-native accounting platform called PocketLedger aimed at this segment, so I have a commercial interest in this market. You should know that when you read my platform views. I would rather name it here than have you discover it later, and the recommendations above are the ones I give clients today.

This one deserves naming separately, because most engagement letters do not address it and it has become a compliance matter rather than a commercial courtesy.

Records must be retained for seven years from the end of the relevant tax period. If an engagement ends, the client still carries that obligation. The provider therefore has to hand back complete accounting data in a format the client can actually use.

My engagement letters now specify this at inception. On termination the complete data set is exported in non-proprietary formats, CSV transaction ledgers, PDF financial statements and a full document archive, delivered within a defined number of days. Agreeing it at the start removes the negotiation from a moment when the relationship may already be strained.

E-invoicing removes the gap between the error and the FTA seeing it. Today the FTA sees aggregate figures at return submission, a gap of months in which errors can be caught internally. Once invoices transmit at issuance, transaction-level accuracy is tested in real time. That requires a pre-transmission review inside the workflow: buyer TRN verified and current, VAT classification correct for the counterparty, numbering sequential, amounts agreeing to the contract.

The second CT cycle creates a consistency obligation. The FTA now holds two years of CT return data for most calendar-year businesses and compares them. Positions from the first return need to carry forward consistently, or the change needs documenting, since an unexplained inconsistency reads as a risk indicator either way. I keep a positions register per client recording what was elected, what policies applied and the basis for each, which runs to two pages and about an hour to build.

IFRS 18 changes financial statement presentation. It replaces IAS 1 for annual periods beginning on or after 1 January 2027, so for a December year end the first statements cover 2027 with 2026 comparatives restated. The changes are largely presentational, new categories and management performance measure disclosures, with modest impact on a straightforward income statement, though property management firms and construction companies need more restructuring. The work belongs in 2026: an impact assessment identifying which line items move, done before the year-end close. The accounting standard itself still runs through Ministerial Decision No. 114 of 2023.

I have not sat through a formal FTA audit examining the quality of accounting records maintained by an outsourced provider and issuing findings on that basis. That specific interaction has not arrived in my practice.

What shaped my approach came through a statutory audit instead. Completing the first audit in one client’s history, over reconstructed financial statements, the fieldwork functioned as a proxy for what an FTA review would look like. Revenue reconciled to tenancy agreements, expenses traced to supplier invoices and payment records, the VAT control account agreed to the filed returns. Every finding the auditor raised was one the FTA could equally have raised.

The finding that changed my operational practice was a category of property management receipts recorded as revenue where the underlying tenancy agreements were incomplete or missing for some units. The receipts were real and the amounts were correct. The documentation trail was not. Resolving it took longer than it should have, because the documents had never been systematically captured at the time.

Since then, every engagement includes a document completeness review before the first month closes. Tenancy agreements matched to rental income, supplier contracts matched to recurring costs, employment contracts matched to payroll entries. The question is whether every significant balance has a source document an inspector could examine without anyone explaining what it relates to.

An internal set of books maintained by the owner can carry ambiguities the owner resolves from memory. An outsourced function’s records have to be self-documenting, because the provider is not there when the FTA arrives. The records have to speak for themselves.

That is the question underneath most first conversations, and it is not about the numbers. It is about whether the person responsible understands enough about how the business works to catch what does not look right, and is engaged enough to actually look.

The anxiety has a specific texture. The accounts are not dramatically wrong, because that would be obvious. It is a background concern that the provider is processing what they are given, closing to a compliance standard, and moving to the next client. Not asking why a margin has been compressing, not flagging a client at ninety days for the second consecutive month.

The worry is not that the provider is doing something wrong. It is that they are doing exactly what was contracted for, and nothing more.

What resolves it is evidence rather than reassurance, and it accumulates through the monthly call. The aged debtors discussion that leads to a collection call recovering money before it becomes a write-off. The margin analysis that catches a pricing error running below cost. The payroll reconciliation that catches a changed direct debit before a third incorrect payment goes out.

Each one is small and individually unremarkable. Enough of them, over enough months, and the question stops arising, because the owner has evidence that the function is paying attention. Book a free consultation with AH Chartered Accountants in Abu Dhabi.