Three currencies, not one
The answer layer trades in three separable currencies. Being ranked buys consideration. Being cited buys credibility and a residual click. Being mentioned buys familiarity that pays off on a later, more specific prompt.
Most teams measure only the third and cheapest of these, because mentions are the easiest to count.
Who actually captures the value
When an engine recommends a brand, it frequently cites a third party as evidence — a review aggregator, a trade publication, a standards body, a public dataset. The recommendation accrues to the brand; the citation accrues to the intermediary.
This creates a strategic choice most brands have not consciously made: compete to be recommended, compete to be the source, or both. The second is cheaper, more durable, and almost entirely uncontested in most categories.
The economics of being a source
Original data is expensive to produce once and nearly free to be cited repeatedly. Explainer content is cheap to produce and almost never cited, because the engine already knows the explanation.
- ›Publish numbers only you can produce, with a stated method.
- ›Give every claim a stable, quotable URL and a date.
- ›Make the units and definitions explicit so an engine can restate them without hedging.
- ›Refresh on a cadence — stale data is dropped faster than it is corrected.
What this means for publishers
Publishers lose the click and keep the attribution. That is a worse deal for advertising models and a better deal for licensing, data and authority models. The publishers doing well in the answer layer are the ones who moved from producing pages to producing referenceable facts.