Pause, hand over or close: different decisions
"Leaving a market" covers three different decisions, and they are often confused.
Pausing stops active marketing and new investment while keeping what is expensive to rebuild: the accounts and their history, the site edition or its useful content, existing customers and the knowledge the pilot produced. Sales may continue through existing channels at a lower level. A pause has a review date.
Handing over moves the market to a different entry mode, such as a distributor or a marketplace, after direct entry did not pay. Marketing continues, but through a partner, and the guide to marketing through a local distributor covers how to keep the brand and its accounts intact under that arrangement.
Closing ends the presence: no new sales, the edition removed or redirected, accounts and lists wound down.
The one-way and two-way doors framework sorts decisions by how reversible they are, and it fits here. A pause is a two-way door: it can be reversed quickly if the conditions change, so it can be decided on lighter evidence. A close is close to a one-way door: deleted accounts, removed content, lost search signals and disappointed partners are slow and costly to rebuild, so it needs stronger evidence. When the evidence is mixed, pause first.
| Option | What stops | What is kept | Reversibility | Suits a market that |
|---|---|---|---|---|
| Pause | Paid media, new campaigns, new local hires or contracts | Accounts, data, useful content, existing customers, the market's records | High | Failed for a fixable reason, or was right but too early |
| Hand over | Direct selling and the brand's own campaigns | The brand's accounts and domain, with the partner given access | Medium | Has demand the business cannot serve directly at a profit |
| Close | Sales, the edition, local accounts and lists | Records and lessons, opt-outs, whatever the business must retain | Low | The evidence says will not work on any affordable model |
Deciding at the gate, and writing down why
The decision belongs at a gate, not in a budget cut. The phase-gate reviews framework asks for criteria agreed before the phase starts and a decision taken against them: expand, hold or stop. If the pilot's proof conditions were written down, the gate reads them; if they were not, the gate is likely to explain the result away.
Before the gate, gather the market's figures on the same definitions as every other market, as the measuring and reporting guide sets them: the agreed key performance indicators (KPIA metric tied directly to an objective.), backend orders and revenue rather than platform conversions alone, and a note of the market's consent setup and calendar so that anomalies are explained. Then write a one-page record of the decision: what the market was meant to prove, what it did prove, why the decision is pause, hand over or close, what would make the business reconsider, and the review date if there is one. That record is the most useful thing a future team will find. The academy exercise on deciding the future of a rowing machine maker's weakest export market practises this decision, whether the market is paused, handed over or closed, and what customers, partners, accounts and the site edition then need. Where the question is whether the market covers its costs, the tutorial on building a contribution view for each market sets its contribution beside the costs that would end and the share of the centre that would stay.
Separate the reasons. A market that failed because the offer did not convert is a different case from one that failed because the localisation was poor, the channel mix was copied from home or the pilot was underfunded. Only the first is a verdict on the market; the others are verdicts on the entry, and they point towards a pause. The international market entry playbook sets this gate up from the start, with proof conditions written before launch and a quarterly decision to continue, adapt, pause or close. To review whether every live market has such a recorded decision, and whether the report behind it compares markets fairly, the multi-market marketing audit scorecard scores both and ranks the fixes.
Telling customers, partners and the media
Existing customers come first. Tell them plainly what changes and what does not: whether they can still order, how existing orders, returns, warranties, subscriptions and support will work, and where to go for help. What the business owes them is a legal and commercial question for advisers in the market; marketing's job is to say it clearly, in the market's language, before they find out another way.
The guide to retiring a feature gives the shape: group customers by how much the change affects them, offer an alternative where there is one, set a phased timeline and change every claim that is no longer true. A market pause or close works the same way, at market scale.
Partners hear it from the business directly and early: retailers, distributors, agencies, media owners and co-marketing partners. Their contracts are questions for the business's advisers; marketing's part is to agree what happens to joint activity, materials in their hands and any accounts they have access to. Local media that covered the launch may ask; a short, honest statement prepared in advance is better than silence, and it should match what customers are told.
Campaigns, budgets and ad accounts
Stop paid campaigns in an order that avoids waste: scheduled and prospecting campaigns first, then remarketing, then anything that still serves existing customers, such as brand search that answers support queries. Remove the market from any shared campaigns, budgets and audiences so spend does not leak into it.
Keep the ad accounts. Pausing campaigns costs nothing; deleting an account loses its history, conversion data and audiences. The media and channels guide recommends separate accounts or clearly separated campaigns per market and keeps ownership with the business; a paused market's account simply sits there with every campaign paused and access reviewed. Remove access for agencies and partners who no longer work on the market, and record who still holds it.
Release the budget at the gate, not quietly over the following months, so the plan for the other markets reflects the decision.
The site edition, search and alternates
The market's site edition holds search signals, links and content that took time to build. What happens to it depends on the decision.
For a pause, keep what is still useful. Product information, guides and support content may still serve customers who own the product and visitors who will become customers on a return. Replace what is no longer true: the checkout, delivery promisesA store's claim about what delivery will cost and when an order will arrive, made on product pages, at checkout, in confirmations and in ads. and prices become a clear notice of how customers in the market can buy now, if at all, and where to get support. Keep the pages in the language and keep their hreflangAnnotation for language/region versions. annotations intact, because the edition is still the right version for that audience. A paused edition still drifts from the source as products change, so the method for keeping localised content in step with the source applies to the pages left live, at least for prices, claims and safety content.
For a close, decide page by page. Google's guidance on redirects, as the site migrations guide states it, is that permanent redirectsA permanent server redirect telling browsers and search engines a URL has moved for good. pass visitors and signals to the destination, and that redirecting many pages to a homepage or a category is treated as a soft 404 for those that do not match, so their signals are lost. Redirect a page only where an equivalent exists that serves the same visitor, such as the same guide in another edition in a language they read. Where there is no equivalent, remove the page clearly with a notice page, take it out of internal links and sitemaps, and accept the loss. Removal is legitimate; silent removal is not.
Then fix the alternates. Google's guidance on localised versions, as the hreflang guide states it, requires annotations to be reciprocal and to point at each page's actual address. When an edition is removed, every remaining edition's hreflang set must drop it, the x-default must still point at a live page, and no annotation may point at a URL that now redirects. Regenerate the sets rather than leaving the old ones in place.
Assets beyond the site, such as marketplace listings, directory entries, partner pages and printed materials, follow the method for withdrawing an asset from every channel: find every place it runs, replace or remove it, confirm, and record.
Email, data and social accounts
Email. Stop promotional email to the market's contacts when sales stop; there is little point offering what people cannot buy. Service messages about orders, support and changes to their account continue as the business's obligations require. The guide to email across countries and languages keeps market and language on every contact and one suppression list, which makes this a filter rather than a project. Any later marketing to these contacts needs a lawful basis and a working opt-out in every message, as that guide and the UK regulator's direct marketing guidance it summarises describe. Keep every opt-out on the suppression list so a return does not mail people who said no.
Data. Customer records are Personal DataAny information relating to an identified or identifiable living person, such as a name, email address, location data or an online identifier.. How long they are kept, and what is deleted when, follows the business's own retention policy and legal advice in the market; this guide sets no retention periodHow long a kind of personal data is kept for a stated purpose before deletion, set with whoever owns data protection and applied in every tool holding a copy.. Keep aggregated performance records, which hold no personal data, for the market file.
Social accounts. A paused market's social account can post a pinned note explaining the situation, answer messages at a reduced level and stay in the business's name. A closed market's account should say so clearly and point to where customers can get help, rather than go silent; deleting it frees the handle for anyone. Remove access for people who no longer work on the market.
What to keep for a return
A market closed today may be the right market in a few years. Keep what would make the second entry faster and less expensive:
- The decision record from the gate, with the proof conditions and the results.
- The research and the market file: customer research, competitor checks, channel maps, costs to serve.
- The localisation assets: style guide, termbase, translation memoryA record, kept by the business, of each source sentence or segment and the translation a reviewer approved for it, so repeated wording comes back the same way., approved copy and creative.
- The accounts: ad, analytics and social accounts kept dormant in the business's name, with access reviewed.
- The domain, renewed, even if it now redirects.
- The contacts of partners, reviewers, agencies and media who did good work.
- The lessons, written as what the business would do differently, not as a narrative of the past.
When the review date comes, the guide to returning to a market you paused or closed reads this record against fresh evidence and sets out what to reuse, what to redo and what to restart.
A yoga mat brand pauses one market (illustrative)
Mivrothe Mats pauses Market B after its pilot gate
Mivrothe Mats, an invented maker of yoga mats, opened Market B as a pilot with a budget for 9 months. At the gate, the cost per new customer is 2.4 times the target and conversion on the localised site is less than half the home rate. Exit interviews with 12 customers and the market's support log show that delivery took longer than the page promised; the offer itself was well received. The team records the decision as a pause, with a review in 12 months once a local fulfilment option exists.
Campaigns stop over one week, prospecting first. The ad and analytics accounts stay, with every campaign paused and the agency's access removed. The 1,800 existing customers receive an email in Market B's language explaining that orders now ship from the home warehouse, that support continues by email, and where the 6 stockists in the market are.
On the site, the checkout and delivery pages are replaced by a notice that lists the stockists; the 14 care and technique guides stay live with their hreflang alternates intact. Promotional email to the market's list stops, and the 230 opt-outs stay on the shared suppression list. The decision record, the research and the style guide go into the market file, ready for the review.
Common mistakes
- Closing when a pause would do. Deleting accounts and content because one pilot failed throws away what a return would need. Pause when the evidence is mixed.
- Deciding in a budget cut rather than at a gate. Without proof conditions and a record, nobody can say later why the market closed.
- Customers finding out from a broken checkout. Tell existing customers first, in their language, what changes and where to get help.
- Deleting ad and social accounts. Their history, audiences and handles are lost. Pause campaigns and keep the accounts in the business's name.
- Redirecting the whole edition to the home page. Mass redirects to an unrelated page lose signals. Redirect only to real equivalents and remove the rest clearly.
- Leaving the removed edition in the hreflang sets. Every remaining edition must drop it, and no annotation may point at a redirecting URL.
- Mailing offers to a closed market's list. People cannot buy; stop promotional email and keep the opt-outs.
- Losing the lessons. Write down what the market taught and keep it where the next team will look.
