How to prove the value of AI in global mobility
A K2 Thinking Paper, November 2026
The research suggests that many organisations aren't there yet. The IBM Institute for Business Value's 2025 CEO Study found that 72% of CEOs consider proprietary organisational data key to unlocking generative AI value, and 68% see integrated enterprise-wide data architecture as critical. Yet in EY's Mobility Reimagined Survey, only 51% of mobility teams trusted the accuracy of their data enough to move to the next phase of AI adoption.
Our new thinking paper, How to prove the value of AI in global mobility, puts data at the heart of the value debate. Its central argument is clear: AI maturity is ultimately constrained by data maturity. However, sophisticated an AI system is, if the data it draws on is inaccurate, incomplete or inconsistent, its output will be unreliable. So will any attempt to prove its value.
The challenge is particularly acute in mobility, where HR, immigration, relocation, tax and payroll data is held by multiple stakeholders and third-party suppliers. Proving strategic value demands even more. Assignment data shows that a move took place. But did the employee settle in successfully? Did the move stay within budget? Did it help the business launch operations in a new market? Without connected post-move, financial and project data, strategic value can't be convincingly proven.
The paper sets out a practical, evidence-led approach to measurement:
- Establishing a robust baseline, combining quantitative and qualitative data, before any AI is introduced
- Forming a testable hypothesis, with defined users, expected benefits, risk measures, a review period and clear criteria for abandoning the initiative
- Using A/B testing to compare the existing and AI-supported processes in parallel
- Tracking leading indicators alongside lagging ones, so you can see how implementation is progressing, not just whether the objective was met
- Calculating net value by weighing what was achieved against the cost of achieving and maintaining it
It also explains why measures such as trustworthiness should never be reduced to a single, universal figure, and why efficiency gains are not automatically financial gains.
For anyone who needs to build an AI business case that stands up to scrutiny, the message is practical: before you measure the value of AI, measure the quality of your data.
Download the full thinking paper
Frequently asked questions
Straight answers to the questions we hear most about workforce mobility.
How do you measure the ROI of AI in global mobility?
+
Start with a clear objective and a robust baseline, then test a defined hypothesis, ideally using A/B testing. Net value is calculated by weighing what was achieved against the cost of achieving it and the ongoing cost of maintaining it.
What does "AI maturity is constrained by data maturity" mean?
+
It means an AI system can only be as reliable as the data it draws on. If mobility data is fragmented, inaccurate or incomplete, AI outputs will be unreliable, however advanced the technology.
What data is needed to prove the strategic value of AI in mobility?
+
Assignment data alone isn't enough. Proving strategic value usually means connecting mobility data with post-move, financial and project data, such as whether the employee settled in, whether the move stayed within budget and whether it supported business goals.
