TR2050 Reward Lab – Zurich, May 2026

The Zurich Reward Lab brought together senior reward leaders across two days of structured sessions to build on the work from previous Labs. The focus was on moving further from debate toward evidence, not ideas about what might work, but structured tests designed to prove it.

The challenge is not knowing what to do. It is proving what works.

Five sessions across two days covered the areas that matter most to the future of reward agenda.

The future of work and technology

Organisations are no longer asking whether AI will change how work gets done. For some, it already has – fundamentally, not incrementally. The group heard from a member organisation that has embedded AI at the core of its business model, and examined what this means for reward and performance design.

Three questions shaped the discussion: whether the job remains the right unit around which to build pay as AI reshapes contribution; whether better tools are producing better decisions or faster versions of existing ones; and where human judgement must remain central in an AI-supported workplace. The competitive advantage, the group concluded, lies not in access to AI tools, which are rapidly becoming a baseline expectation, but in proprietary data, the quality of people using it, and the organisational will to redesign rather than automate.

Performance management and incentive design

The group examined whether individual ratings are fit for purpose, and reached a clear view. The evidence points not to ratings as the problem, but to the link between ratings and pay. That link weakens feedback quality, drives rating inflation, and creates defensive behaviour.

Decoupling ratings from pay, and directing differentiated reward more deliberately toward a defined group, emerged as the most promising direction. The group was candid about the preconditions required for this to work, and about how many organisations are not yet there.

The question of team-based versus individual incentive design was also covered in depth. The group examined what the evidence actually says about when team incentives work, where they don’t, and what organisational conditions need to be in place before they are worth testing.

Merit pay and analytical evidence

A member organisation shared original research testing a widely held assumption about what drives sales performance. The findings were unexpected, and significant enough to reshape thinking about where merit budgets have the most impact and why.

The most striking finding concerned not pay itself, but something most organisations treat as a separate question. The group agreed a clear next step: a structured pilot with a defined geography, a measurable outcome variable, and a two-year window to assess impact.

Benchmarking and reward technology

A live demonstration of a new approach to compensation benchmarking addressed structural problems that the current market has not solved – around data ownership, timeliness, transparency, and trust. The group tested it in real time, submitting data and receiving results without any intermediary involved.

A specific first use case was agreed and is ready to move forward: salary budgeting in high-inflation markets, where multinationals need real-time peer data that official statistics cannot reliably provide.

AI-enabled tools, skills-based pay, and test-and-learn

A scan of over 300 reward technology vendors found that AI capability in the market is more widespread than most practitioners assume, but that capability alone is not the right criterion for adoption. What matters is whether a tool genuinely improves the job, and whether it is backed by strong post-sale support.

On skills-based pay, the group concluded it is not yet ready for universal implementation, and may not need to be. The data problem precedes the pay problem. And two live tests already in progress were shared with the group, with findings to be presented at the next Lab.

What comes next

Across five sessions, four findings stand out:

• Leadership continuity matters more than merit increases in driving sales performance
• Decoupling ratings from pay is gaining real traction
• Decentralised benchmarking is technically feasible with a first use case agreed
• Structured, hypothesis-driven testing is becoming a genuine competitive advantage for the organisations doing it.

Pilots and tests are now underway across these areas. The next Reward Lab takes place in New York in September.