The IT Onboarding Process When the New Starter Is in Another Country

TL;DR

• IT onboarding is everything that has to happen before somebody can work on day one, and in a distributed company most of it has a lead time measured in weeks rather than hours.
• If new starters come to an office and collect a laptop from a desk, you do not need this. You need a checklist and a spare machine.
• The binding constraint is almost never configuration. It is getting a physical device into a country, which customs can extend without warning.
• Work backwards from the start date, not forwards from the offer. The order of operations is what makes day one work.
• The single highest-value change is ordering hardware at offer acceptance rather than at contract signature, which typically buys you ten days.
• Measure the proportion of starters productive on day one. It is the only number that captures whether the process worked.

The Laptop That Arrived on Day Nine

A company hired its first engineer in a new country. Everything was done properly and in the obvious order: contract signed, then IT notified, then a laptop ordered from the supplier they used at home, then shipped.

The machine cleared the border on day seven and reached the employee on day nine. For the first week and a half the new engineer had a personal laptop, no access to anything containing company code, and a manager apologising daily. They spent the time reading public documentation and attending calls on a phone.

Nothing in that chain was a mistake in isolation. The contract came first because that is how hiring works. IT was notified on signature because that is the trigger in the HRIS. The laptop came from the usual supplier because that is the approved route. The failure was that a sequence designed when every starter was ten minutes from a storeroom had been applied unchanged to somebody 6,000 miles away, where the same sequence has a three-week critical path. Nobody had drawn the path.

Recommended Tools

Best tools for Device Management

B BambooHR HR management system L Lattice Performance management D Deel Global payroll & EOR

View all recommended tools

So the useful exercise is not a better checklist. It is working out which step is binding and starting it earlier.

When You Don't Actually Need to Change Anything

When the manual way is genuinely fine

Starters come to an office, there is a cupboard with spare machines, and somebody hands one over on the first morning. The process is a checklist and a person, and it works because the lead time on every step is minutes. Nothing in this article applies, and the trigger to revisit it is the first hire who will not be in the building.

When friction starts appearing

A manager asks, before the start date, whether the laptop will arrive in time. That question is the signal, because it means somebody has already been burned and is now carrying anxiety the process should carry instead. The useful response is to find out what the honest answer is rather than to reassure.

When it becomes a liability

The point at which a new starter's first experience of the company is a week of waiting. The cost is not the lost week, which most managers can absorb. It is that you have told somebody, in their first fortnight, that the company is disorganised, and they will repeat that to the next candidate you want to hire from their network.

The edge case that forces it

A cohort, or a first hire in a market with difficult import arrangements. A cohort converts a series of individually tolerable delays into one visible failure, and a difficult market can extend the physical lead time to a point where no amount of internal efficiency helps.

Five Questions People Ask First

"What actually takes the longest?" The device, essentially always. Account creation, access provisioning and software installation are measured in hours and can be done the day before. Procuring, shipping and clearing a laptop into another country is measured in days to weeks, and it is the only step with genuine external dependency.

"Can we just ship from head office?" You can, and for one or two devices it is often the slowest and most expensive option available. A machine sent from your home country is an international shipment of goods with a commercial value, so it attracts the full customs process. Buying locally in the destination market frequently costs less and arrives sooner, even where the unit price is higher.

"When should we order?" At offer acceptance rather than at contract signature. These are typically seven to fourteen days apart, which is most of the gap you are trying to close, and the risk is a small number of accepted offers that fall through leaving you with a machine. For most companies that machine goes to the next starter, which makes the risk a timing cost rather than a loss.

"Who owns this end to end?" Usually nobody, which is the real finding. HR owns the hire, IT owns the configuration, procurement or finance owns the purchase, and the handoffs between them are where the days disappear. One named owner for the sequence, with visibility of all three, is worth more than any tooling.

"Does the new starter need to do anything?" Two things, and both should be asked at offer stage rather than in week one: confirm a delivery address, and confirm somebody will be there to receive a parcel. A failed delivery adds three to five days and is the most avoidable delay in the whole process.

Working Backwards From Day One

The order matters more than the list, so here it is as a timeline rather than a checklist. Day zero is the start date.

Day minus 21, at offer acceptance. Confirm the delivery address and a receiving window. Order the device, from the destination market where possible. This is the only step that cannot be compressed later, which is why it goes first and why it goes before contract signature.

Day minus 14. Create the identity record. Not the accounts, the record: the person exists in the HRIS with a start date, a manager, a role and a location, because everything downstream derives from it. Most provisioning failures trace to this being done late or incompletely.

Day minus 7. Provision accounts and group memberships from the role, and have the device enrolled in management so it configures itself on first boot. Confirm the shipment is moving and, if it is not, escalate now rather than on day minus one, because a week of notice is the difference between a workaround and an incident.

Day minus 3. Send the starter their own information: what is arriving, when, what to do when they open it, and a named person with a phone number if anything goes wrong. In their own language if that is not your working language.

Day minus 1. Verify the device was delivered and that accounts resolve. This is a five-minute check and it is the one that catches the failure early enough to warn somebody.

Day zero. The starter switches the machine on, signs in, and the configuration applies itself. If the preceding steps happened, this step is uneventful, which is the entire goal.

Day plus 5. Confirm the asset record shows the device assigned to the person, with the serial and the delivery address. Onboarding is not complete until the register says so, and this is the step that is always skipped.

The Timeline as a Table

The two rows people move are the first and the last. Moving the order later makes the start date a gamble. Dropping the day-five check makes the device unidentifiable in eighteen months. Both feel harmless at the time, which is exactly why they need to be written down as steps with owners rather than left as good intentions.

The Four Routes to Getting Hardware There

Ship from your home country

Right for a market with uncomplicated imports where you already have a working lane and a courier who knows your paperwork.

It fails on time and on cost in most other cases, because the device is a commercial import and goes through the full process. It also fails on repair later, since warranty cover is commonly tied to the country of purchase.

Buy locally in the destination market

Usually the fastest route, often cheaper all-in despite a higher unit price, and it keeps warranty and support in the country where the person actually is.

It fails on consistency, because the exact model may not be available and you will accumulate specification variants. It also requires either a local entity, a card that works, or somebody who can buy on your behalf.

Use a global procurement platform

A vendor who holds or sources stock in the destination market, configures it and delivers domestically.

Right when you hire in several countries regularly and the administrative load of local buying has become the constraint. It fails on cost at low volume, where the service charge is large relative to one laptop, and coverage depth varies by market in ways the headline country count hides.

Have the employee buy and expense

Fastest of all and the one most companies resist, correctly in most cases.

It is right for an urgent start in a market you have no presence in. It fails on everything afterwards: the device is not yours in any clean sense, enrolment is a negotiation, the specification will not match, and recovering it at the end is a conversation rather than a process. Where it is used, treat it as a deliberate exception with a plan to replace the machine at the next refresh.

What the Starter Should Receive Before Day One

The day-minus-three message is listed above and not specified, which makes it the kind of step that gets reduced to a generic welcome email. Here is what it should contain and why each part earns its place.

What is arriving and when. The device, the courier, the tracking reference and a date range. A person who knows a parcel is coming on Thursday can arrange to be there; a person who does not know will be out.

What to do when it arrives. Two lines: switch it on, connect to wifi, sign in with the address we have sent you, and leave it alone for twenty minutes while it configures. Most enrolment failures are somebody intervening during setup because nothing appeared to be happening.

What will not work on day one, stated plainly. There is always something: a system that needs manual approval, a tool whose licence is ordered monthly, access that depends on a manager being available. Saying so in advance converts a defect into an expectation.

Who to contact, by name, with a route that does not require the new account. A phone number or a personal messaging handle. A new starter locked out of setup cannot email you from an account they are trying to activate, and this is the single most common way day one becomes day two.

In their working language if that is not your main one. This is short, consequential, read once, and read by somebody with no context to fall back on.

Keep it to one screen. The instinct is to attach the handbook, the org chart and the benefits guide, all of which are useful in week two and all of which reduce the chance that the five things above are read.

Send it from a named person rather than a system address, for the same reason the contact detail matters: a starter with one small question needs somewhere to send it, and a no-reply sender converts a thirty-second clarification into a day of waiting. The message is also the first thing many new starters receive from anybody at the company other than their recruiter, which is worth remembering when deciding how it reads.

How to Choose: Five Questions Before You Talk to Any Vendor

What is your actual critical path, in days? Measure it from your last three international starters rather than estimating. Most teams find the device path is two to three times what they assumed, and the number changes the conversation immediately.

How many markets, and how often? One hire a year in each of four countries is a different problem from twenty a year in one. Platforms are priced and built for recurrence.

Do you need the device back? If yes, the procurement decision and the recovery decision are the same decision, because a device bought locally by an employee is considerably harder to recover than one delivered by a platform that knows its serial and its address.

What is your tolerance for specification drift? Local buying produces variety. Decide in advance whether you are specifying a processor class and memory, which survives market differences, or exact models, which will not.

Who will own the sequence? Ask this of yourself rather than the vendor. A platform removes the logistics and does not remove the need for somebody to own the order of operations.

The Options

Figures below were read from each vendor's own pricing page, most recently 9 October 2026. Almost nothing in the physical logistics half of this category publishes a price.

Buying direct in each market

Best for: one or two hires a year per country, where the administrative cost is lower than a platform fee.

Why it works: fastest, keeps warranty local, no vendor relationship to manage. For a company hiring occasionally this is frequently the right answer and the one least often considered.

Where it struggles: needs a payment method and somebody willing to do it, produces specification drift, and gives you no help at all with recovery later.

Global procurement and lifecycle platforms

Workwize, Deel IT, Firstbase, GroWrk and allwhere all source, configure and deliver hardware internationally, and none of them publishes a price. They differ most on country coverage depth rather than on features, so shortlist against your actual markets and treat headline country counts sceptically. Expect to compare quotes.

RemoAsset

Disclosure: RemoAsset is owned by the same people who publish HROpsLab. It appears here because it competes in this category and is assessed against the same criteria as everything else on this page, with its limitations stated in the same detail.

Best for: companies where onboarding and offboarding are the same problem, meaning you need the device back as reliably as you need it delivered.

Why companies choose it: the delivery creates the asset record, so the serial, recipient and address are known from the start rather than typed afterwards, and the same platform triggers the return at offboarding. For a fleet that churns, this closes the loop that usually leaks.

Where it struggles: it publishes no price and requires a demo. Coverage depth varies by market like everything else here, so check your specific countries. And it is not a certified disposal vendor, so the end of life is a separate arrangement.

Freshservice

Best for: teams who want the onboarding workflow itself, meaning the tasks, approvals and assignments, rather than the hardware logistics.

Why companies choose it: tiered at $19, $49 and $99 per agent per month, it covers the internal choreography well, with onboarding templates that fire the right tasks at the right people.

Where it struggles: it orchestrates and does not ship. The binding constraint in this article is physical, and no workflow tool moves a laptop across a border.

The Comparison

The Decision Table

What Goes Wrong at Each Handoff

The timeline assumes the steps connect. They mostly do not, and the gaps are predictable enough to name.

Hiring to the sequence owner. The offer is accepted and nobody tells the person who orders hardware, because the HRIS trigger fires on signature rather than acceptance. This is the single most common failure and it costs the entire gap between the two events. Fix it by making offer acceptance a notification, which is a configuration change rather than a project.

Sequence owner to procurement. The order needs a payment method, an approval or a supplier account in a country you have not bought in before. Discovering that at the point of ordering adds days. Find out which of your markets you can actually buy in before you have a starter waiting.

Procurement to the courier. The address is incomplete, lacks a phone number, or nobody is there. Couriers in many markets make one attempt and then hold the parcel, and reclaiming it is slower than the original delivery. The confirmation at offer stage prevents nearly all of this.

Courier to customs. Outside your control and not outside your planning. What you can do is know the typical clearance time for your route, build it into the day-minus-21 figure rather than treating it as an exception, and have the paperwork complete so a hold is not caused by you.

Delivery to the register. The device arrives, the starter is working, everybody moves on, and nothing records which serial went to which person. This is the quiet one, and it produces an unidentifiable device at offboarding. It is also the handoff with no natural pressure behind it, because every other gap has somebody waiting and complaining, while this one has nobody at all. That is precisely why it needs a named step on a named day rather than an intention to update the record once things calm down.

So the practical version of all this is a single owner watching five joins rather than five teams each watching their own step. Each team genuinely completes its own task promptly, and the elapsed time is still three weeks, because elapsed time is nobody's metric when the work is divided this way.

Access Provisioning, and Why It Is the Easy Half

This article has argued that the device is the binding constraint, which risks implying that accounts and access are trivial. They are easy in elapsed time and they are where the second-week complaints come from, so they deserve a short treatment of their own.

Provision from the role, not from a copy of a colleague. The common shortcut is "give them the same as Sarah", which propagates Sarah's accumulated exceptions to every subsequent hire in that team. Within two years the role has access nobody chose and nobody can justify, and unpicking it is far harder than defining it once.

Separate day-one access from week-two access deliberately. A new starter needs identity, email, calendar, chat, the document store and their primary tool. Everything else can follow, and front-loading the full set means a longer approval chain gating the things they actually need on the first morning.

Name the approvals that are human. Some access requires a person to agree, and that person has a calendar. If an approval cannot be obtained in advance, it belongs in the day-minus-three message as something that will not work on day one, rather than being discovered by the starter.

Set the leaver path at the same time as the joiner path. Provisioning from a role means deprovisioning from a role, and the teams that do the first well and the second badly are the ones where access removal is a manual list that drifts. The moment to define removal is while you are defining grant, not eighteen months later during an access review, because at that point you are reconstructing intent from a list of permissions rather than reading a definition somebody wrote deliberately.

Write down what the role does not get. A role definition that lists only grants is ambiguous about everything else, so each new request becomes a judgement made under time pressure by whoever is asked. Naming the common things a role explicitly does not receive, and who can approve an exception, turns those into a short conversation rather than a precedent set by accident.

The reason this stays the easy half is timing: every one of these steps can be done the day before, repeatedly, at no cost. The device cannot. That asymmetry is the whole argument of this article and it is worth stating plainly rather than leaving implied.

Measuring It

Four numbers, and the first is the only one most companies lack.

Proportion productive on day one. Able to sign in, reach the systems their role needs and do actual work. This is the outcome measure and everything else is diagnostic. Ask the manager at the end of week one rather than inferring it from ticket data.

Device lead time, from order to in-hand. Measured per market, because the average across markets is meaningless. The spread between your fastest and slowest market is usually the most actionable fact in the whole exercise.

Days between offer acceptance and device order. This is the controllable gap and it is almost always larger than anybody believes. It is also the cheapest thing to fix, because closing it requires a trigger change rather than a budget.

Asset record completeness at day five. The proportion of new starters whose device appears correctly in the register within a week. This is the step that gets skipped, and skipping it is where the offboarding problem starts eighteen months later.

The Spare Machine Question

Everything above optimises a sequence. There is a blunter instrument that most distributed companies dismiss too quickly, which is holding stock in the markets where you hire.

The arithmetic is better than it looks. A spare laptop in the country where your next hire will be converts a three-week critical path into a domestic shipment measured in days, and it removes the customs variable entirely. The cost is the capital tied up in a machine plus its depreciation while it waits, which for a device that waits two months is a modest figure against a week of a new hire's time.

It only works where you hire repeatedly. One spare in a market you hire into once every two years is a laptop that will be three years old and obsolete before it is issued. The rule of thumb that holds up is to keep stock in markets where you expect at least three or four hires a year, and to run the sequence properly everywhere else.

The holding location is the hard part, not the decision. Options are a lifecycle platform with regional warehousing, an employee or office willing to hold a box, or a local supplier who will keep stock allocated to you. The second is more common than anybody admits and works until that person leaves.

Set a maximum holding period. A machine that has waited six months should be issued to somebody, anybody, rather than continuing to wait for a perfect match, because depreciation does not pause. Pair the stock decision with a rule that says stock older than two quarters gets deployed at the next opportunity regardless of role fit.

And treat returned devices as the natural source of spares rather than buying new ones. A laptop recovered from a leaver in the same market, wiped and held locally, is the cheapest possible answer to the next hire in that market, and it closes the loop between the two halves of the lifecycle. It also changes the economics of recovery itself, because a returned machine that becomes next quarter's spare has an obvious value, whereas one shipped to a central store and never reissued does not, which is often why recovery struggles to get funded in the first place.

What Getting This Wrong Costs

The direct cost is paid time with no output, and for a week of a senior hire it is a real number, though rarely large enough on its own to force a change.

The second cost is the impression, which is harder to reverse than it is to avoid. A new starter forms a view of the company's competence almost entirely from administrative interactions in the first fortnight, because there is nothing else to judge yet. "My laptop took nine days" is a specific, memorable and repeatable story, and the people telling it are the ones you were hoping would refer their former colleagues.

The third is the compounding one. A device that arrives late was usually also recorded late or not at all, because the person handling it was focused on getting it there rather than on the register. So the onboarding failure plants an asset record failure, which surfaces at offboarding as a device nobody can identify. The two problems look unrelated and are the same problem eighteen months apart.

The fourth cost lands on the manager, and it is the one that quietly changes hiring behaviour. A manager who has had two starters sit idle for a week begins to hedge: they delay the start date, they tell the next candidate to expect a slow first fortnight, or they push to hire in the market where the laptop arrives quickly rather than the market with the better candidate. None of that is recorded as an onboarding problem. It appears as a hiring preference, and it is a procurement constraint shaping where the company employs people.

So the question worth asking before the next international hire is not whether the process is documented. It is how many days the physical path takes, and whether anybody has started it yet.

When You're Ready to Move Beyond Ordering on Signature

Nearly every company triggers hardware on contract signature, for a sound reason: that is the moment the hire is certain, and nobody wants to buy a laptop for somebody who does not join. The logic is sound and the cost of it is invisible, which is why it survives.

What makes it stop working is distance. When the gap between signature and start is ten days and the device path is three days, the trigger is fine. When the device path is fifteen days, the same trigger guarantees a late arrival for every international hire, regardless of how efficient everybody downstream is.

The change is small and it is mostly about who is allowed to spend. Move the order trigger to offer acceptance, accept that a small number of machines will be bought for people who do not start, and plan for those to go to the next hire rather than treating them as waste. Then buy in the destination market where you can, name one owner for the whole sequence, and confirm the asset record at day five so the loop closes. None of that needs a platform, and after a quarter of running it you will know from your own numbers whether one would help.

Frequently Asked Questions

HROpsLab takes no vendor money and publishes no paid placements, which is why this page says "publishes no price" for every logistics platform rather than estimating.