Enterprise A-Z: eXperiments
Let Small Tests Do Some of the Arguing
You can spend months in theoretical debates about what “might” work. Legal is nervous, finance is cautious, operations is stretched. Everyone wants more proof before committing – but the proof can only come from actually trying something.
That’s where experiments come in.
Why experiments matter
Experiments are how you move from opinion to evidence.
They shrink the stakes so people can say yes to something now.
They turn abstract risks and opportunities into concrete learning.
They give you real numbers and stories to take into bigger rooms.
Without experiments, you’re stuck trading slides and beliefs. With them, you can say, “Here’s what we saw in three months in this one plant / product / market.”
What experiments actually are
An experiment is not a vague pilot that drifts on indefinitely. It’s a deliberately small test with:
A clear question – “If we do X, will we see Y?”
A defined scope – one site, one product line, one supplier, one market.
A time‑bound window – usually a few months.
Simple success criteria – what you’ll measure to decide whether to scale, adapt or stop.
Your job is to design experiments that feel safe enough for cautious colleagues to try, but sharp enough to produce learning the enterprise can use.
How to start
Pick a leverage point, not a corner. Choose an area where a small test could have outsized implications if it works – a key product, a meaningful supplier, a visible site – rather than the place where it’s easiest to get permission but no one will care about the outcome.
Define “small, clear, fast.” With a willing sponsor, write down on one page: what exactly you’re testing, where, for how long, what you’ll measure, and how you’ll decide what happens next. Make the resource ask and the risk profile proportionate to the size of the test.
Use the results to change the conversation. When the experiment ends, resist the urge to declare victory and move on. Harvest what you learned – numbers, operational realities, human reactions – and take that back into the rooms where people were hesitant. “Here’s what we actually saw” lands very differently from “here’s what we believe.”
You don’t need dozens of experiments running at once. You need a few well‑chosen tests that help your organisation experience, in a contained way, what a different future could look and feel like – and then decide, with more confidence, how far and how fast to go.
