69% of mobility leaders want AI-powered scenario modeling.

Written for CFOs, VPs of Finance, and FP&A leaders who own the mobility line item.

When we surveyed 50+ global mobility, HR, and finance leaders for The AI Inflection Point, we asked which AI capability would deliver the most value to their program. Scenario modeling for assignment planning came first, ahead of real-time compliance monitoring, policy matching, and automated cost projections.

For a finance audience, that result should land differently than it does for anyone else in the room. Scenario modeling is not an emerging capability. It is the foundational discipline of financial planning. You already run scenarios on revenue, headcount, capex, FX exposure, and debt structure. You have been doing it for decades.

So when the function that spends six figures per assignment names scenario modeling as its number one unmet need, the honest reading is this: one of the most expensive decisions your company makes has been running without the analytical rigor you apply everywhere else.

The last six-figure decision nobody models

An international assignment is a multi-year financial commitment. Base salary, housing, cost-of-living adjustment, schooling, tax equalization, gross-up, shipping, home leave, and repatriation stack into a number that frequently rivals a mid-size capital project.

A capital project of that size gets a business case, a sensitivity table, and a committee. An assignment of comparable value typically gets a cost estimate built in a spreadsheet, under time pressure, with one set of assumptions, by someone who will not remember six months later which assumptions they used.

When your cost data sits in Excel, your immigration timelines sit in a vendor portal, and your tax data sits in a different vendor portal, there is no version of that spreadsheet that constitutes a source of truth.

“

That is the gap the 69% are pointing at. Not a shortage of expertise. A shortage of infrastructure.

What scenario modeling actually looks like in practice

The abstract version of this capability is easy to nod along to and hard to act on. Here is the concrete version.

Take a real decision: a senior engineer moving from the US to Germany for 24 months. The question your business partner asks is "what will it cost?" The question you should be answering is "what are our options, and what does each one cost?"

Each scenario returns total employer cost, cost per year, and the components driving the difference. Built by hand, this is three days of work and two spreadsheets, and realistically you build two of the six. Built in a system designed for it, you clone the baseline, change one variable, and re-run.

The output shift matters more than the time savings. You stop delivering a single number that carries an implicit and unearned claim to accuracy. You start delivering a range with named drivers. The conversation with the business changes from "is this estimate right?" to "which trade-off do we want?" That is a conversation finance is built to lead.

Three mechanics separate a real modeling capability from a calculator with a nicer interface.

It was not a close race for second. Almost nobody is fighting the technology. They are fighting for the money to use it. Which means the tool mobility teams want most is precisely the tool that would help them justify the budget they don't have.

The way out of that loop is to model the cost of not modeling.

Build the case backwards

Don't start with what modeling costs. Start with what the absence of it costs. Every input below is something finance already owns, beginning with the one you can size in about a minute.

Take your assignments and relocations per year, your average total employer cost per move, and your typical estimate-to-actual variance. Program spend multiplied by your own variance is your annual budget exposure sitting outside your forecast. This is the number that shows up as a surprise in Q4. Then take analyst hours per cost estimate: the time spent rebuilding estimates is often roughly 0.2 FTE, spent producing one scenario per decision instead of six.

That first number is the one to walk into the budget conversation with. The rest of the case fills in around it.

Compliance exposure

84% of leaders are not fully confident they know where their assignees are or whether they are compliant. Permanent establishment risk, unremitted payroll withholding, and visa lapses are contingent liabilities you currently cannot size. 85% said an AI system that automatically monitors employee locations and compliance events would be valuable, with 62% calling it very valuable. The need is not in dispute. The funding is.

Attrition and experience

Only 15% of leaders rated the experience they provide to internationally assigned employees as excellent. 62% called it good but with friction. One declined assignment or one early repatriation costs more than most modeling tools do in a year.

Decision quality

The hardest cost to quantify and the largest. Every assignment structured on a single unexamined assumption is a decision made without alternatives. You cannot recover that spread retroactively.

Finance teams don't fund tools. They fund outcomes. Frame this as forecast accuracy and risk reduction rather than as a software purchase, and it stops competing with software purchases for budget.

You don't need clean data to start

The most common objection is the data problem, and it is real. But waiting for a fully integrated mobility data estate before running your first scenario model is a decision to never run one.

The practical move is to start where your data is already decent. For most organizations that is cost estimation, precisely because finance already touches it. Compensation figures, allowance schedules, and tax rates are the cleanest inputs in the mobility stack. That makes cost scenario modeling the lowest-friction entry point available to you, and it happens to be the capability with the highest demand.

Pilot it against a real, upcoming decision rather than a hypothetical. Show your leadership the old way and the new way side by side. Tangible beats theoretical.

The signal in the 69%

77% of organizations are already using or piloting AI in their mobility programs. 84% expect to increase that investment. Not a single respondent said AI was off their radar. Whatever your position on the technology broadly, the mobility function is not waiting.

What they are asking for is not a chatbot or a novelty. They are asking for the ability to answer, with evidence, a question finance has always asked: what are the options, what does each cost, and what are we exposed to if we're wrong?

That request deserves a yes.

Talk to a mobility expert

Horizon by Topia is an AI-native platform purpose-built for global mobility, combining scenario-based cost modeling, natural-language policy management, and automated tax and immigration compliance in a single system.

Survey data throughout is drawn from The AI Inflection Point: How Artificial Intelligence is Reshaping Global Mobility, based on responses from 50+ global mobility, HR, and finance professionals. Read the full report.