A customer sees an ad, searches for your brand and buys. The advertising report records a conversion. What it cannot establish from that journey alone is whether the customer would have bought anyway.
Attribution assigns credit to recorded touchpoints using chosen rules. It helps describe the journeys you can observe. Incrementality asks a different question: what additional outcome did the advertising cause?
A controlled experiment compares comparable groups with different exposure to advertising. With a sound design, it can help estimate the extra effect. Platform lift studies are one option where available; they are not available or suitable for every account.
Simply turning a campaign off and comparing this week with last week is weaker evidence. Seasonality, promotions, stock and other marketing may also have changed.
Not every business has the scale or resources for a convincing study immediately. Be clear about that limitation. Use attribution for its useful job, and resist presenting attributed revenue as proof that every sale was additional.
