Actions

Every other view tells you what’s happening. This one tells you what to do about it. Actions is the playbook — a ranked list of concrete moves, each one tied to the buyer questions it would change, each one tested before it reaches you.

The five kinds of work

Actions group by the kind of change they ask of you. The grouping mirrors the four signal states from Concepts: your standing on each question points to a kind of work, and the work lives here.

Update Existing Content — you already have pages the engine is picking up on, and these edits help them land. Restructuring for clarity, filling in missing context, backing a claim with a source. This is the considered-not-cited fix, and it’s usually the fastest.

Create New Content — the engine is answering questions about you where you have no page. These are the net-new pages worth building to fill the gap.

Counter Narratives — the engine is saying something about you that’s wrong or unflattering. These actions target the specific claims driving that take so you can reshape it.

Influence Third Party — reviews, directories, and press are shaping what the engine says about you. These are chances to improve what those sources say or earn citations from new ones.

Leverage Recency — updated dates, current stats, recent references. Recent content tends to land better, and these actions lean on that.

Quick Win and Strategic

Each action carries an effort tag. Quick Win is a small, contained change you could ship this week. Strategic is a larger investment. Sort by Effort to line up the quick wins, or by Simulated Impact to lead with the moves that shift the most.

Tested before you invest

This is the part that separates the playbook from a list of suggestions. Every action was run back against your audit’s buyer questions to predict whether it would actually move the answer engine in your favor. The result shows as a lift figure — the predicted change in how the engine treats you once the action is live. You see the expected payoff before you spend a dollar or an hour, so you’re putting effort where it’s already shown it will pay.

You can test your own version too. If you’d write the change differently from the suggestion, edit it and re-run the prediction to see how your wording scores — see Simulations.

Opening an action

Click an action to open it. Pages names what to touch or build. Why explains the problem it solves. Suggested Approach is the concrete how. Simulated Changes shows the before-and-after on the affected questions — the engine’s answer as it stands now, and as it’s predicted to read once you’ve done the work. The impacted questions link back to Prompt detail so you can read the full context behind any one of them.