How to Leave the Middle East Without Closing It for a Decade

Nobody Uses the Word

The decision is almost never called an exit.

It is a budget decision. A headcount decision. A reallocation of resource towards a region that is converting faster. The territory is folded into a neighbouring profit centre, the two remaining people are given something else to do, and the travel line comes out of next year’s plan. At no point in that sequence does anybody say the company is leaving the Middle East, and at no point is anybody lying.

The market experiences it as leaving, because the market only ever sees behaviour.

The first article in this series ended on the real cost of a badly run entry — the sentence said in a leadership meeting a year later, we tried that region, it didn’t work, which converts a strategy failure into a market verdict and closes the file. This article starts where that one stopped. Once the file is closed, something still has to be done with what is inside it.

A vendor that stops is holding things that belong to other people. Support obligations on systems that are still running. Stock and unbilled effort a partner has already paid for. Commitments made to a government client on the record. And a reputation among perhaps thirty people who will still be in the same seats in five years’ time. Hand those over deliberately and the market stays open. Walk away from them and it closes — not by decree, but because the same names decide the next one.

Not a Hand Over. A Leftover.

I should be straight about what I have and have not done.

I have never closed the Middle East. What I have been asked to do, more than once, is shrink it — and the discipline of abandoning nothing while you do that is the same discipline a clean exit requires. Two earlier pieces in this series describe those contractions from the inside.

The one full exit I have worked through end to end was Europe’s, and I was on the receiving side of it.

It was hasty and effectively overnight. Measured against what I have heard about since, it was handled better than most: every purchase order was honoured, pending installations were completed, and teams were merged quickly enough that no commitment lapsed outright. It could still have been done a great deal better.

What it looked like in practice was this. The first problem was service, because systems were live and the people who supported them had gone. Calls were routed to whoever could take them — teams in the United States, the Middle East and India covering remotely, immediately, before anything else was settled. Then partners, one at a time. Then the CRM, line by line.

That last part is the one nobody budgets for. Contacts, open orders, commitments and pipeline sat partly in the system and partly in the heads of people who were no longer employed. Reconstructing that record was laborious, and it was the single most valuable thing we did — everything afterwards depended on it.

Six months in, the one salesperson who remained and I closed two large orders in a market the company had just walked out of. Five months after that he left too, and someone from another team took it on. Then we did the only thing that reliably works: we went and saw people.

Not every partner came back. The ones who mattered did — and the ones who mattered turned out to be, almost exactly, the ones with something at stake. A demo unit on their floor. An open order. Live pipeline. We rebuilt around them and then built past them: a new partnership in the United Kingdom, a demo centre in central London co-funded with that partner, a relaunch primed and ready to run.

Then COVID arrived and everything stopped.

I do not offer that as a success, because it never got the chance to be one. I offer it as evidence of something narrower and more useful: a partner was willing to put his own money into a brand that had walked out of his market two years earlier. You only get that if the walking out was done in a particular way.

Communication and physical presence are what tell a market it has not been abandoned. There is no substitute for either, and no memo performs the function of a visit.

What You Are Actually Holding

Most companies discover at this point that the region was never a line in the forecast. It is an inventory, and the majority of it is not on the balance sheet.

Open orders, some paid for or part-paid. Demo units sitting in partners’ showrooms. Spares and tools. Support contracts with time left to run. Enquiries that were weeks away from becoming orders, and the pipeline behind them. An office, a trade licence, a set of registrations, and the process of relinquishing each one. And the people who built it, who are about to become the market’s most credible source on how your company behaves under pressure.

Two of those categories behave completely differently, and separating them is the first useful act.

The legal obligations are finite and expensive to ignore: taxes due, compliance filings, labour obligations, open support contracts. They do not lapse because the region did, and in this part of the world several of them attach to named individuals rather than to an entity.

The reputational exposure is the larger hit, because it compounds. News travels here faster than any announcement, competitors amplify it, and what they amplify is not always accurate. There is exactly one defence and it is not a press release. It is a direct conversation with each of the people and partners who matter, held proactively — which in practice means held before they hear it from somebody else.

Then the part almost nobody plans for. A good deal of that inventory can be converted rather than written off.

We did all four. It does not turn a loss into a win. It converts part of the loss into an asset, which at that point is the only move on the board.

The Installed Base Does Not Care That You Left

Systems keep running. Warranties keep running. Somebody has to answer the phone.

The published answer is usually adequate on paper — a support address, a service desk, stated hours of coverage. We had a 24/7 desk and a support inbox, and neither of them was the thing that mattered. What mattered was that customers had an escalation point with a name attached: one person they could reach when the process did not work. The desk is the contract. The name is the reassurance, and it is by some distance the cheaper of the two.

Handover has a natural order of preference. If a distributor is already in place, it is theirs — that is usually what happens and usually the right answer. Failing that, someone competent outside the market can hold the relationship, but only if there is a capable third party inside it doing the work. Remote support without local hands is a promise, not a service.

Spares and firmware are where customers actually get stranded. If a vendor is going permanently, some of that is unavoidable. What is avoidable is leaving without releasing what you still can — software keys, remaining stock, documentation — so that the customer and the local partner can sustain the system between them.

Government and critical installations are not the same conversation at all. The obligation is heavier, the consequences travel further, and in this region they can reach the individuals who signed as readily as the company they signed for.

What holding that line looked like is already on the record here: installations under warranty maintained and projects mid-delivery completed straight through a restructuring, with the region still renewing service contracts on that work today; and a cloud platform wound down in two stages, with private instances stood up for the handful of accounts whose workflow genuinely required it. We lost nobody. Neither exercise was cheap and neither was heroic. They were decisions taken in the right order.

The Partner Is the One Who Gets Asked

Your partner introduced you to their customers. When you go, those customers do not ring you. They ring the partner. Whatever you do to that partner is therefore done in public, in front of the only audience whose opinion will still matter if you ever come back.

Most of what you owe is already written down. Stock, margin on deals in flight, notice periods, termination mechanics — the partnership agreement covers it, and you honour it. That part is administration, and companies generally get it right.

What no agreement covers is the individual who represented you. He carried your name into his own accounts and staked his credibility on it. When you leave, he loses the relationship and the trust that came with it, and there is no clause that compensates him for either.

The difference between an ending a partner respects and one he does not is almost entirely sequence. Plan the transition, tell him early, tell him honestly, and he will be disappointed and fine. Spring it on him and leave him holding the ball with nothing to say to his own customers, and he will not be fine — and he will say so, for years. There is precedent for that in this series: a distribution relationship that ended in acrimony left a lasting mistrust, inside the vendor, of the entire distribution model, long after everyone involved had moved on. Endings do not merely close relationships. They change behaviour.

On commercial agency law, a caveat first. This is commercial guidance and not a legal opinion, jurisdictions differ, and the position has moved in recent years — the protections bite less absolutely than their reputation suggests, and the sensible course is proper advice on the specific case rather than expensive caution about the general fear. Then the warning, which is the part that actually governs the outcome.

A Government Client Is a Record, Not a Relationship

It is in fact both, and the two halves are handled separately.

The formal half is exactly as formal as the contract requires, and no more. Where notice is owed, give it properly and in writing. Where none is owed, there is no benefit in manufacturing a filing nobody asked for.

The relationship half is not optional and it is not a letter. The people who hold the relationship get a meeting, and in that meeting they are told plainly how this works from here — who supports the installation, which partner or contractor is taking it on, and who they call. What you are protecting in that room is their position, not yours. Nobody in a government post wants to learn from a competitor that the system they personally specified no longer has a manufacturer standing behind it.

Then the quiet part, which is how most companies actually lose their standing. Prequalification, vendor registration and agent registration all run on renewal cycles. They are rarely withdrawn. They lapse — because a renewal notice arrives at an address nobody is reading, and nobody answers it. Not by decision. By absence. And six months of not answering costs considerably more to undo than several years of keeping the file current.

Keeping that door open is unglamorous and cheap. An invitation to the headquarters event you are running anyway. A meeting at the trade show they already attend. Enough contact that when the requirement comes back, you are a name somebody can put forward without first having to explain who you are.

The Market Is Small and It Remembers

The line you hear is that a vendor who exits badly finds the same three names blocking them five years later. Put like that it sounds ominous and slightly unserious. The real mechanism is more ordinary than that, and more consequential.

It is not a blacklist. It is promotion.

The engineer on my first project in this region became a section head. A junior IT support staffer at one of our Saudi installations turned up years later as the consultant on a mega-project we went on to win. That is not a Gulf phenomenon, it is a global one — but the community here is small enough that you watch it happen: the same faces at the same trade shows year after year, with a different badge each time.

Which cuts both ways, and is why the honest version of this argument is less frightening than the folklore. No exit is permanent. If you have something of value, the door reopens. If you left well, the return is quicker and somebody on the inside will vouch for you. If you did not, the return is still available — it simply costs what it cost the first time, which is the full grind of a first hire in a market that owes you nothing. That is a real price, and it is described in detail elsewhere in this series. But it is a price, not a wall.

What Is Worth Keeping Alive

If there is any prospect of returning, a dormant position is an asset that can be maintained deliberately. The mistake companies make is treating every element of it as the same kind of decision. There are two categories, and only one of them is a cost-benefit question.

Commercial items are the cost-benefit category. Keep what would be expensive to rebuild, drop what would not, and let the circumstances decide. In practice the single most valuable item on that list is also close to the cheapest: relationships. A quarterly visit. Invitations to events you are holding regardless. A scheduled hour at the two trade shows everybody attends. And, if it can be justified at all, a skeletal presence in region.

That is not a slogan. A quarterly visit cadence is a working model with its own economics and its own failure modes, and it is the subject of the first article in this series.

Legal and people items are not a cost-benefit question. Whether to hold an entity dormant depends on the outlook and on the reason for leaving. But closing one properly is a process in its own right, it deserves its own legal advice, and it is never the place to economise. Loose ends do not stay with the company that created them. They attach to shareholders, directors and authorised signatories personally, and they surface at the least convenient moment available — which is usually the moment somebody is trying to register something new.

The minimum viable presence, in the end, is smaller than most people assume: an open line of communication, and a reason to appear four times a year.

What I kept alive personally through the worst of it was narrower still. Service. Support. And answering enquiries — including the ones arriving from markets we were no longer supposed to be covering.

When Leaving Is the Right Answer

Everything above assumes that staying, or returning, is the goal. Sometimes it is not, and an adviser who cannot say so out loud is selling something.

Withdrawal is the correct decision more often than the market-entry industry admits. What matters is that it is reached in a defensible order — because most exits are argued backwards. The conclusion arrives from head office and the reasons are assembled afterwards to fit it.

Underneath all of it sits one distinction, and getting it wrong is precisely what produces the sentence in the leadership meeting.

A market that has rejected you and an operating model that has failed look identical from head office. They are not the same thing. Conditions change, technology changes, and a company genuinely can stop fitting a market it once fitted — that is a real finding and it deserves respect rather than a turnaround plan. But where the fit is there and the operation is not working, the problem is structural, and the answer is a different structure, not a different continent.

If the fit is not there and there is no path to it, decide sooner rather than later. If the fit is there and the economics are not, what you have is a structure problem wearing a market’s clothes.

What I Got Wrong

Three things, and the first is less a mistake than a limit.

I have never designed an exit. I have only ever arrived in the middle of one. Every version of this I have lived through was already underway by the time it reached me — the decision taken elsewhere, the people already leaving, the sequence in this article assembled afterwards out of what worked and what did not. So read the checklist as what I wish had been in place, not as something I have ever run from the top. There is an uncomfortable corollary: the companies that most need it are the least likely to be able to use it, because the decision to leave is almost never taken by the person who will have to carry it out.

We left that market with nobody in it for five months. The last salesperson went, and it was five months before anyone was assigned to take over. There were reasons. The company itself did not yet know what shape it was going to be in, and you cannot commit a person to a market while that is unresolved. It was still five months. Five months in which every partner and every customer had nothing to look at except an absence, at precisely the point when each of them was deciding privately whether the brand still existed. Whatever the constraint, that period cost considerably more than the salary it saved.

And I cannot tell you why most of those partners did not come back. I went back more than once. Then I stopped, and the specific reasons have gone from my memory entirely. At the time each was an individual disappointment with its own explanation — someone who had moved on, a line that had been replaced, a call that was not returned. Collectively they were most of the channel we had spent years building.

That is the argument of this entire article, and I am describing it from inside the mistake rather than from outside it. The partners who did come back were the ones with something at stake. I would like to claim I engineered that. I did not — I worked it out afterwards, which is the only reason it is in here.

If You Are Considering This

The sequence matters more than any individual decision in it, because almost every irreversible loss in an exit happens in the first fortnight, before anybody has thought about the year after.

Done in that order, what remains is a dormant position that one person with a plan can reopen. Done in the wrong order, or not at all, what remains is a market that has quietly reclassified you — and thirty people who will still be there when you want to come back.

The cost of leaving badly is not paid at the exit. It is paid at the return, by somebody who was not in the room when the decision was made.

Connektions MEA works with vendors already in the region whose position is not performing — including the ones deciding whether to stay. That work is described under Regional Reset.