We are hired when a decision has become expensive to postpone. The work is to establish what is true, test the narrow question, and write down a position the board can hold to.
Four weeks inside your analytics, contracts, and cost lines. Most of what a publisher believes about its AI exposure is inherited from a conference talk. We replace that with your own numbers before anyone argues about what to do.
Narrow pilots, each answering a single question, each cheap enough to abandon. Every pilot carries a stopping rule written before it starts: the result that would end it, agreed by the people who would otherwise defend it.
One written position, with the reasoning and the numbers left visible, so it can be argued with rather than believed. Abandoned options are recorded and priced in the same document, which is what stops them returning next year.
Fixed fee, one written position at the end. If we do not think we can help, we will say so on the first call.