When Your Company Rewards Certainty But You Need Experiments

I was in a boardroom last month watching a product director present three months of user research. Solid work. Clear insights. Reasonable hypotheses about what might work.

The CEO's first question? "What's the ROI guarantee?"

This is the innovation paradox that kills most transformation efforts before they start. Companies say they want to be more experimental, more agile, more customer-centric. Then they ask for business cases with three-year revenue projections for ideas that haven't been tested yet.

You can't experiment your way to breakthrough innovation while demanding certainty at every step. The two approaches are fundamentally incompatible.

But here's what I've learned working with brands like NatWest and Disney: you don't need to change the entire company culture to start experimenting. You need to change how experiments are framed, funded, and measured.

Reframe experiments as risk reduction, not risk taking

The language you use matters more than you think. In certainty-driven cultures, "experiment" sounds like playing around with the budget. "Risk mitigation through rapid testing" sounds responsible.

I watched a team at American Express get funding for customer journey experiments by positioning them as "validation sprints to reduce implementation risk." Same activity, different frame. The CFO approved it because it sounded like due diligence.

When you present an experiment, lead with what you're trying to avoid, not what you're trying to achieve. "This two-week test will help us avoid spending six months building something customers don't want" lands differently than "Let's try this cool new feature."

Start with operational experiments, not product experiments

Product experiments feel risky because they touch revenue directly. Process experiments feel safe because they're about efficiency.

We helped ING build their experimentation muscle by starting with internal operations. How might we reduce meeting time by 20%? What if we tested asynchronous decision-making for non-urgent choices? Can we experiment with different onboarding approaches for new hires?

These experiments built the methodology and mindset without triggering the corporate antibodies that attack anything touching customer experience or revenue streams.

Once teams saw that small tests could produce measurable improvements in how work gets done, they started asking what else they could test. The experimentation mindset spread organically.

Design experiments to produce certainty, not eliminate it

Here's the counterintuitive part: good experiments actually increase certainty. They give you data instead of opinions. Evidence instead of assumptions.

Frame your experiments as certainty-generating activities. "After this test, we'll know definitively whether customers prefer option A or option B" is more compelling than "Let's see what happens."

At NASA, we structured experiments around decision points the organisation needed to reach anyway. Instead of asking for permission to experiment, we asked for permission to gather data before making required decisions. Much easier conversation.

The trick is designing experiments that answer specific questions executives already have, not random curiosities your team wants to explore.

Three concrete steps you can take this week

First: Identify one decision your team needs to make in the next month. Frame a simple test that would give you data to make that decision more confidently. Present it as research, not experimentation.

Second: Find an internal process that frustrates people and propose a two-week pilot to test an alternative approach. Keep it low-stakes and operational. Measure something concrete like time saved or satisfaction scores.

Third: Create a simple experiment brief template for your team. Include sections for "What decision does this help us make?" and "What risk does this reduce?" Use it for every test proposal.

The goal is not to trick your organisation into becoming experimental. It's to demonstrate that systematic testing reduces uncertainty and improves outcomes. Once people see experiments producing better decisions with less risk, the culture shift follows naturally.

You don't change culture by arguing for different values. You change it by proving different methods work better.

You can't experiment your way to innovation while demanding certainty at every step, but you can use experiments to create the certainty your culture craves.