Buying laptops in one country is a supplier relationship. Buying in ten is a different job, and the thing that breaks first is never the thing teams prepare for. It isn’t cost. It’s that nobody in the company can answer a simple question about what hardware exists, in which country, and who is currently holding it.

TL;DR

  • Somewhere around eight countries, country-by-country improvisation stops scaling and starts losing devices.
  • The first failure is the register, not the budget. Nobody knows what exists where.
  • Standardise the specification before the supplier. One build list beats ten supplier relationships.
  • Decide buy-locally versus ship-in per country rather than as one global policy.
  • Retrieval is the half that gets forgotten, and it is where the cost and the loss sit.
  • Country coverage claims differ widely between vendors, so check your actual list rather than headline totals.

What breaks, and in what order

It fails in a predictable sequence, and knowing the sequence tells you what to fix first.

The register goes first. Each country buys from a different supplier, nobody records consistently, and within a year the spreadsheet is directionally right and specifically wrong. Somebody asks how many machines exist in Poland and the honest answer is a range.

Then the specification drifts. Local buying produces local stock, so your Spanish team runs a different model from your Dutch team, with a different keyboard, a different warranty and a different support path. Support becomes ten support problems.

Then retrieval collapses. This is the expensive one. Nobody set up a route for getting machines back in markets where they were bought ad hoc, so leavers keep their laptops by default and the loss shows up as a procurement line rather than as a loss.

Cost comes last and arrives as a symptom. By the time anybody is complaining about spend, the other three have been broken for a while.

Standardise the build before the supplier

Teams usually attack this by consolidating vendors. Start one step earlier.

Agree two or three builds. Express each one as a minimum specification rather than a named model: processor class, memory, storage, and an operating system you support. A specification survives a model being discontinued and a supplier being swapped, and a named model does not. It also makes every quote you receive comparable, which is most of the work.

Then publish it. A hiring manager in any country should be able to read what their new starter gets without asking anybody, and the moment that’s true the number of one-off decisions drops sharply.

What to do:

  • Write two or three builds as minimum specifications, not model numbers.
  • Publish them with an exception route and a named approver.
  • Record every device against a person and a country, from the first order.
  • Decide buy-or-ship per country, in writing, and revisit it annually.

Buy locally or ship in, country by country

There’s no single right answer, which is why a global policy fails. The decision belongs per market and turns on three things.

Can you import at all? Commercial goods entering a country need a party able to act as importer of record. Where you have no entity that’s usually not going to be you, and buying inside the country removes the whole question before it is asked.

How many people are there? One person in a market rarely justifies a supplier relationship. Twelve people probably does justify one, and the economics flip somewhere in between those two numbers depending on how often you hire there.

What does support look like? A machine bought locally usually carries warranty you can actually use. One shipped in from elsewhere may not carry anything usable, and that difference only ever becomes visible at the exact moment something breaks.

A company at eleven countries

A 260-person company hired into 11 markets over three years with no central process. Each country lead bought what they could get.

The audit that finally forced a change found 7 distinct laptop models in use, three different warranty arrangements, and 23 devices whose location nobody could confirm. None of that was anybody’s fault individually, because every single purchase had been reasonable at the time, and that’s exactly how this failure works: it is an accumulation of sensible decisions nobody was ever asked to reconcile.

Fixing it took two standard builds, one register and a per-country buy-or-ship decision. The 23 missing devices were never recovered.

Retrieval, which is the half that gets dropped

Worth its own section, because almost every multi-country procurement project covers outbound and ignores the return.

Getting a laptop to somebody is a solved problem in most markets if you’re willing to pay. Getting one back from somebody who has already left, in a country where you have no office and no relationship, is a different exercise, and the recovery rate difference between a planned route and an improvised one is large enough to dominate the whole cost picture.

So ask about it first rather than last. If a vendor quote covers delivery only, you have a quote for half the problem.

Comparing platforms on coverage

Headline country totals are the least useful number on any vendor page. Check your eleven, not their hundred.

RemoAsset is our own product. It covers procurement and retrieval across 80+ countries with each device tracked from issue through return, from $1,499 a year on the Subscription tier for companies with 50 to 5,000+ employees, including a 60 per cent reduction on retrieval charges.

Three limits. It has no network discovery, so it knows about devices it handled rather than the ones your country leads bought themselves, which is a real gap when you’re cleaning up exactly that situation. It has no software licence management. And only the Subscription tier publishes a figure, with pay-as-you-go quoted on request.

On coverage, Rippling IT states more than 130 countries for inventory and shipping, which is a larger number than ours, and it publishes no price. Workwize, Firstbase, GroWrk, allwhere, Esevel and Deel IT publish nothing either.

Checklist

Checklist:

  • Two or three published builds, written as minimum specifications.
  • One register with a device, a person and a country on every row.
  • A buy-or-ship decision recorded per country, not globally.
  • Your actual country list checked against each vendor individually.
  • Retrieval scope confirmed in writing before any contract.
  • An annual review, because supply routes and headcount both move.

Final Thoughts

Multi-country procurement rarely fails loudly. It degrades, through a series of reasonable local decisions that nobody reconciles until an audit asks a question somebody can’t answer.

The fix is unglamorous and it works: standardise the build, keep one register, decide buy-or-ship per market, and treat retrieval as half the problem rather than an afterthought. Do those four and the cost question tends to answer itself.

Frequently asked questions

At what point does ad hoc procurement stop working?

Around eight countries. Though it depends more on how many markets hold only one or two people than on the headline total. The first symptom isn’t cost, it is that nobody can say what exists where, because each country bought from a different supplier and recorded it differently. By the time spend is the complaint, the register has usually been wrong for a year.

Should we standardise on one laptop model globally?

Standardise on a specification rather than a model. A minimum specification covering processor class, memory, storage and a supported operating system survives discontinued models and swapped suppliers, where a named model does not. It also makes quotes from different countries directly comparable, which is most of the benefit.

Is it better to buy locally or ship from one place?

Decide per country rather than globally. Where you have no entity you often can’t act as importer of record, which makes local purchase the practical route. Where you have a dozen people in a market, a local supplier relationship starts to pay for itself. One person in a market usually doesn’t justify a supplier relationship of its own, however much it might feel tidier to have one in every single market you operate in.

Why does retrieval matter more than delivery?

Because delivery is solved if you’ll pay for it, and retrieval is where devices actually disappear. Getting a machine back from somebody who has already left, in a country with no office and no relationship, is a different exercise, and the gap between a planned route and an improvised one dominates the real cost. Ask about it before delivery, not after.

How should we compare country coverage between vendors?

By your list, never theirs. Headline totals are close to meaningless because the long tail of any vendor’s coverage is where service quality varies most. Send every vendor the same specific country list and ask them to answer yes or no per market, including whether retrieval is covered there as well as delivery.

What do we do about devices nobody can locate?

Accept the losses. Concentrate instead on stopping the next batch from going the same way. Logistics platforms generally know about the devices they handled rather than ones bought locally before they arrived, so a clean-up relies on asking people directly and recording what comes back. Set a cut-off date, record what you can confirm, and start the new register from there.