Shared service centres in this region were built on a simple proposition: move repetitive finance and administration work to wherever competent people cost least, hold a service level agreement over it, and bank the difference. For fifteen years that was the entire business case. Two things have changed it in the last eighteen months, and neither is technological. Saudi Arabia's regional headquarters requirement took effect at the start of this year, and corporate tax arrived in the Emirates last June. Between them they have converted a cost decision into a market access and substance decision.
A shared service centre used to be a decision about where labour is cheap. It is now a decision about where you are allowed to sell
That reframing is uncomfortable for anyone holding a five-year business case built on wage differentials.
Three forces, none of them about efficiency
Market access. Government entities in the Kingdom no longer contract with companies that lack a regional headquarters presence there. For groups with meaningful public sector revenue, the location question now has a revenue answer attached, and it outranks any operating cost comparison. Tax substance. Once a jurisdiction has corporate tax, intragroup service charges stop being an internal accounting convenience and become a transfer pricing position. A hub that invoices eleven entities for services needs people with authority, decisions genuinely taken there, and documentation that says so. The arbitrage narrowed. Automation removes the high-volume, low-judgement work that the wage differential was earned on. What remains is judgement-heavy, more senior, and priced far more similarly across locations than processing ever was.
Substance is the design principle, not a compliance afterthought
The old model tolerated a thin hub: a modest team executing instructions issued elsewhere. That design now fails two tests simultaneously. It fails the tax test, because a service charge has to reflect functions performed, assets used and risks assumed at the place performing them. And it fails the operational test, because the remaining work is exactly the work that requires someone empowered to decide rather than to process. The practical translation: put real decision rights in the hub, name the people who hold them, minute the decisions, and make sure the organisation chart and the transfer pricing file describe the same company. Groups that get audited on this rarely fail on the arithmetic of the cost-plus margin. They fail because nobody in the hub could explain what they decide.
What belongs in the hub and what does not
Centralise transactional accounts payable and receivable, payroll processing, master data maintenance, management reporting, treasury operations and the exception desks that sit over them. Leave local: statutory filing and tax submission, because rules, languages and authority relationships are local and the penalty for being wrong is local too; credit decisions on local customers, where the relevant knowledge is relationship knowledge; anything requiring a licensed professional in that jurisdiction; and anything gated by a physical document that has to be presented in person. The common mistake is centralising statutory work because it looks like accounting. It is not accounting. It is a filing obligation attached to a legal entity and a person.
The intercompany agreement is what gets examined
Write the service agreements before go-live, not in the following year's tax review. Define the services, the allocation basis, the mark-up and the benefit each recipient entity actually receives. Then make sure the recharges that run each month match what the agreement says, because the mismatch between the document and the ledger is what an examination finds first. This is unglamorous and it is the single highest-return hour of work in the entire programme.
Do not move a process you have not fixed
Transition amplifies whatever it inherits. A process that works because three experienced people quietly correct it does not survive relocation, and the failure appears six weeks later as an unexplained reconciliation difference. Document the process as it actually runs, fix the obvious breaks, then move it. Overlap the old and new teams through at least one full month-end, and budget for the receiving team to be slower for a quarter. Every transition plan that assumes otherwise is rewritten in month two.
Practical Guidance for Regional Hub Design
- Start from market access obligations, then optimise for cost.
- Put genuine decision rights in the hub and minute them.
- Write intercompany agreements before go-live, and match the ledger to them.
- Keep statutory filing and tax local, whatever the org chart suggests.
- Fix processes before relocating them, not after.
- Overlap teams through a full month-end, at minimum.
- Check localisation quotas in the receiving jurisdiction during design.
- Model the tax position of the hub itself, not just its operating cost.
The Regional Angle
Saudi Arabia has made the most decisive move, and the detail matters more than the headline. The regional headquarters programme pairs a hard commercial condition — eligibility to contract with government entities, in a market where government and government-related spending is the dominant buyer — with a substantial tax incentive package for qualifying headquarters activities running for decades. The conditions are real: strategic direction and management functions actually performed in the Kingdom, minimum staffing including senior executives resident there, and a defined list of eligible activities. Groups treating this as a registration exercise will find the eligible activity list narrower than hoped, and groups treating it as an operational relocation will find the incentives materially better than the accounting suggests. Either way the decision belongs to the commercial side of the business rather than to finance operations, because the variable being optimised is addressable revenue. The Emirates presents a subtler question, which is free zone versus mainland for the hub entity itself. Qualifying free zone status can deliver a zero rate on qualifying income, and intragroup services sit in a part of the rules that requires careful reading rather than optimistic assumption — the treatment depends on what the services are, who receives them and where the recipients sit. Getting it wrong is not a minor adjustment; it can put the entity's entire qualifying status at risk for the period and the periods following. Take a written position from a tax adviser before choosing the jurisdiction of incorporation, not after the lease is signed, and make sure whoever designs the recharge model has read that position. The operational attractions of a free zone — ownership, setup speed, visa quotas — have quietly become secondary to a tax question that did not exist two years ago. The rest of the Gulf is more varied than the shorthand suggests, and the variation is operational rather than strategic. Each jurisdiction runs its own nationalisation targets with its own definitions and penalties, so a hub headcount plan that works in one country can be non-compliant in another with identical numbers. Banking access, account opening timelines and the ease of moving cash between group entities differ considerably by country and by bank. Labour rules on notice, end-of-service and contract types vary enough to change the cost of flexing the team. And for groups that need scale in genuinely high-volume processing, Egypt remains the region's deepest and cheapest labour market with a strong finance graduate pipeline — which argues for a two-tier design more often than a single location: judgement and decision rights in a Gulf hub where substance and market access demand them, volume processing where the labour economics still work.
The objection worth taking seriously
The strongest objection is that the whole concept is dated. Shared service centres were an answer to a 1990s problem — expensive local processing teams duplicated across countries — and both remote working and automation have dissolved that problem. If the work can be done by people anywhere and increasingly by software, building a physical hub with a lease, a floor plan and a relocation budget looks like solving a 2024 question with a 2004 instrument. For pure processing that is largely correct, and any business case still resting on consolidating data entry into one building deserves the scepticism. What has changed is that the reasons for physical presence are no longer efficiency reasons. Substance requirements mean the location of people and decisions has tax consequences that no distributed model addresses. Market access rules attach eligibility to a registered, staffed presence in a specific country. Both are legal requirements rather than operational preferences, and neither is satisfied by a well-run remote team. There is also a quieter operational argument: the work that remains after automation is judgement work, and judgement work benefits from proximity to the business it serves in ways that transaction processing never did. The correct conclusion is not to abandon the hub concept but to stop justifying it with wage differentials. Build it where the obligations require, staff it with people who decide rather than process, and let the cost case be a secondary benefit rather than the headline.
Common Questions
Does a hub need to be in the same country as most of our revenue?
Not necessarily, but where market access rules tie contracting eligibility to local presence, revenue exposure should drive the decision ahead of operating cost.
Can one hub serve entities in multiple countries?
Yes, and that is the normal design. The constraint is that each recipient entity needs a defensible service agreement and a demonstrable benefit.
How many people make a hub substantive?
There is no universal number. What matters is whether identifiable people there hold real authority and can evidence the decisions they took.
What should we expect over the next twelve months?
Expect more groups to establish or upgrade a Saudi presence during the year, and expect competition for senior regional finance talent to tighten as a result. Expect the first corporate tax filings in the Emirates to surface transfer pricing questions about intragroup charges that have run unexamined for years. Expect other jurisdictions in the region to respond with incentive packages of their own. And expect hub business cases written before 2023 to be reopened, because the variables they optimised are no longer the variables that decide.
Regional Hub Design — we start from your market access and substance obligations, then build a hub that survives both the tax review and the first month-end.