Most experiments come at a cost.
A cost in attention.
A cost in patience.
Sometimes even a cost in trust.

What if you could remove that?

Synthetic audiences offer a way to test, learn, and filter ideas before they ever touch a real customer. On paper, it sounds like the perfect fix—no wasted impressions, no survey fatigue, no erosion of goodwill.

But here’s the question that keeps me up at night:

👉 Does removing the customer from the testing equation make experimentation more ethical—or less?

Because yes, synthetic can spare customers from endless asks.
But if experiments only run in “synthetic labs,” do we risk designing for a reflection of reality instead of reality itself?

Where’s the line between protecting customers and drifting too far from them?