Two hundred servers. Fifteen countries. One go-live date.
The instinct is to treat a multi-country IT deployment as one shipment repeated fifteen times. It is not, and the projects that run late are usually the ones planned that way.
Carra Globe sells into this problem, so we have a commercial interest in how you answer it. What follows is written to be usable at planning stage whether or not you ever speak to us, including the parts where the answer is that your customer should import.
What changes in a multi-country IT deployment
The work does not multiply, it changes shape. Freight scales predictably. Importer eligibility, product approvals and registrations do not, because each is decided separately in every destination.
One shipment into one country is a logistics question. The same hardware into fifteen is fifteen independent regulatory questions that happen to share a purchase order.
| Scales with volume | Decided separately per country |
|---|---|
| Freight cost and consolidation | Who is eligible to act as importer |
| Warehousing and staging | Type approval and product certification |
| Insurance value | Producer and recycling registrations |
| Installation resource | Import licensing and permits |
Read the right-hand column. Nothing on it gets easier because you are doing it fifteen times, and none of it can be answered once and applied everywhere.
Which means the planning work is front-loaded rather than repeated. Five things decide whether the dates hold.
The five things that decide the timeline
- Which markets you can lawfully import into. Some require a locally established importer. That is a feasibility question, not a paperwork one, and it can remove a country from the plan entirely. Our IOR by country index sets out the position.
- Which SKUs trigger approvals. Wireless, mains-powered and encryption-capable equipment are the usual triggers. The consequence is disproportionate: a single SKU without the approval it needs can hold the consignment it travels in, so a rack of compliant switches waits on one access point. The usual fix is to split that SKU onto its own shipment, which costs a separate customs entry, extra freight and a second delivery window your customer has to accept.
- The longest-pole country. One market with a multi-month approval route sets the date for any site that must go live simultaneously.
- Who is named as importer in each destination. It does not have to be the same party everywhere, and often cannot be.
- Where the equipment stages. A single consolidation point simplifies freight and complicates customs, because goods that enter a country and leave again raise their own treatment questions.
Three staging models, and they are not interchangeable
If you are planning a consolidation hub, the customs treatment is the decision, not the warehouse.
- Temporary admission suits goods with a defined return. It generally needs security or a guarantee, and you have to evidence the re-export. The internationally recognised framework sits in Specific Annex G of the Revised Kyoto Convention, which sets out the conditions and the time limit for re-exportation.
- Bonded movement keeps the goods under customs control in transit, and depends on the countries involved recognising the transit arrangement.
- Import then re-export is the simplest to arrange and the most expensive, because it triggers a duty and tax event at the hub that you then reclaim or absorb.
Decide which before choosing the hub, not after. The right answer changes which country the hub should be in.
These are international frameworks rather than a single global rule. The World Customs Organization sets out the underlying standards for temporary admission, transit and processing procedures, but each country implements them in its own legislation, so confirm the treatment available in the specific market before committing stock to a hub.
Sequence by lead time, not by priority
Of those five, one is a decision rather than a discovery, and it is the one most often made backwards. It separates rollouts that land from rollouts that slip.
Most programmes are sequenced by business priority. Biggest site first, or the region with the loudest stakeholder. That works until the market with the longest approval route turns out to be scheduled near the end.
Start the slowest markets first, even where they matter least. Approvals and registrations run on national timelines that are largely outside your control. Freight is generally the more flexible of the two to accelerate. Sequencing the easy countries first feels like progress and simply moves the problem to the end of the programme, where there is no time left to solve it.
Which raises the obvious question: how do you know which markets are slow before you have researched all fifteen?
You can predict most of it structurally. A market tends to be slow when three things stack: telecom or radio type approval handled by a national regulator, a separate product standards or conformity regime, and an importer registration that must be obtained before anything else can start. Where local testing is required rather than foreign test reports being accepted, add more.
A market tends to be fast when approvals rely on recognised international test reports, there is no separate importer registration to obtain, and conformity is self-declared rather than certified by a regulator.
Use that as a screen, not an answer. Actual lead times move, and a market that was quick last year may not be this year. Check the current position per country on our IOR by country pages rather than working from a figure quoted in an article.
The practical output is a two-column list before anything is booked. It does not need to be sophisticated.
| Destination | Longest lead item | Owner |
|---|---|---|
| Market A | Radio type approval, local testing required | Third-party importer |
| Market B | Importer registration before anything else | Third-party importer |
| Market C | Product standards certification | Customer |
| Market D | None. Freight only | Own entity |
Sort it by the middle column and you have your programme sequence. The order of work follows that list, not the org chart.
One importer or several?
A common assumption is that a global rollout needs a single importing entity. It usually cannot. Importer eligibility is determined by each destination, and some markets require local establishment or other specific qualifications.
What you can have is a single point of coordination with different entities named per country. In practice a fifteen-country rollout often ends up mixed.
- Your own entity in the handful of markets where you already have one. Administrative rather than structural.
- The customer where they are registered and willing, which is usually the cheapest route and worth asking about before assuming otherwise.
- A third-party importer in the markets where neither of you qualifies, or where the customer has declined the role.
The order to work through it is the same in every market. Do you have a local entity there? If not, is the customer registered and willing? If not, is a third party permitted to act in that country? Three questions, and the first yes is your answer.
Mixing them is normal and not a sign of a badly planned programme. What matters is that the classification, the declared values and the documentation stay consistent across every entry, because the same SKU is being declared fifteen times and the numbers need to agree.
One commercial consequence worth planning for: a mixed model produces uneven landed costs across the programme. Markets where the customer imports carry no third-party fee but shift duty and compliance exposure to them. Markets where you appoint a provider carry a fee but a predictable landed price. If you are quoting a single global figure, that variation has to be averaged somewhere, and it is better done deliberately than discovered per invoice.
Where you have no local presence, our page on importing without a local legal entity covers the routes.
How Carra Globe helps
Multi-country IT deployment is a core part of what we do, usually reaching us with the destination list already fixed and a go-live date attached.
- Screening the destination list first, so a market that will not work is known before it is in the plan rather than after.
- Acting as importer of record across 175+ countries, including for data centre equipment, with one point of contact rather than fifteen.
- Handling DDP, staging and warehousing and white glove delivery to the rack, so the last mile is not a separate project in every market.
- Flagging the longest-pole country early, which is the single most useful thing anyone can tell you at planning stage.
Underneath it sits global trade compliance, and our case studies cover comparable deployments, including a data centre import from the USA into Cambodia.
Send the destination list and the SKU list. We will tell you which markets are straightforward, which need a structure, and which one is going to set your date.
Screen your destination list →
Frequently asked questions
Can one company import into every country in a rollout?
Not necessarily. Importer eligibility is set nationally, and several markets require a locally established entity, so the named importer often differs by destination.
A single coordinating provider is achievable. A single named importer everywhere usually is not.
Should I ship to all countries at once?
Where the programme depends on a common go-live date, start the markets with the longest approval or registration lead times first, even where they are lower business priority.
Freight is generally more flexible to accelerate than a national approval or registration process, which is why sequencing by business priority is a common reason rollouts slip at the end.
Does consolidating shipments simplify a multi-country deployment?
It simplifies freight and can complicate customs. Goods that enter one country and then leave again raise their own treatment questions, which vary by jurisdiction.
Worth deciding the staging model early, since it affects both cost and the customs position.
What usually delays a global IT rollout?
Product approvals and importer eligibility rather than freight. Both are decided per country, neither scales with volume, and both are discovered late when the destination list is not screened first.
Our guide to calculating landed cost covers the commercial side of the same planning question.
Disclaimer: This guide is for informational purposes only and does not constitute legal or customs advice. Importer eligibility, product approval requirements, staging treatments and import procedures are set nationally, vary by destination and by product, and change over time, so none of the guidance above should be applied to a specific market without checking the current position there. This article reflects publicly available information as at 9 September 2026. Always confirm the position for your equipment and destinations with a qualified adviser or the relevant customs authority.