Skip to main content
The Prefer Action Center is the board inside Prefer that turns tracking and audit data into ranked AEO tasks. Each card shows an impact score, the prompts and engines affected, the evidence behind it and a step-by-step plan, and Prefer measures citation lift once the work is done. Last updated 3 October 2026

​
The short version

The Prefer Action Center is where tracking data becomes a to-do list: every card is one fix, ranked by expected citation gain for the effort. Monitor tells you where you stand in ChatGPT, Gemini, Perplexity, Google AI Overviews and AI Mode. The Action Center answers the next question, which is what to do about it. It sits under Act in the app, next to Content Studio. Cards arrive in the Suggested column after each tracking run or audit. You accept the ones worth doing, dismiss the rest, and drag accepted cards through To do, In progress, In review and Done. A List view is there when you would rather sort and filter than drag.

​
The seven kinds of action

Every Prefer action belongs to one of seven categories, and the card shows where it came from. Three categories are generated live from tracking data. Four come from your latest site audit, with best-practice templates filling in when no audit exists.

​
What is on a card

Open a card and you get three sections. The claim says what the action should achieve. The proof shows the evidence, for example a competitor’s citation count, a blocked crawler or a missing source, plus the exact prompts and engines where the gap showed up. The plan lists numbered steps with copy-ready assets such as outreach email templates, robots.txt or schema diffs, checklists and fact sheets. The impact score is not a black box: it is Reach, Winnability and Confidence divided by Effort. Effort is marked low, medium or high, so you can batch quick wins. Each card also carries an owner label (agent, assisted or human). The Action Center docs are clear that this label is a filter, not an automation trigger. In the current release every action is done by a person.

​
How Prefer checks that a fix worked

When you accept an action, Prefer records a baseline of citations or mentions from the current tracking run. After the change is live and a later run completes, the Done card shows three numbers:
  1. Before: citations or mentions when you accepted the action.
  2. After: the same count on the most recent run.
  3. Lift: the absolute and relative change.
The count uses the same citation logic as the Sources and Citations view, so numbers match across the app. Outcomes stay empty until a later run completes, and dismissed actions record no baseline. This before-and-after check is the point: a fix only counts when the citation number moves.

​
A first session on the board

  1. Open Act, then Action Center, and sort the Suggested column by impact.
  2. Filter by effort and pick three low-effort Technical or Schema cards.
  3. Accept them, copy the robots.txt or JSON-LD diff from each plan, and ship the change.
  4. Move each card to In review once the change is live, then to Done.
  5. Check the lift on each card after the next tracking run.

​
Plans and how to try it

Prefer’s pricing page lists the prioritized action engine on Growth, Scale and Enterprise. Starter does not include it. The free trial switches on every premium feature, so you can work real cards before paying. To see a starter list of fixes for your own brand first, run the free AI visibility audit, which ends with a short ranked action plan.

​
Frequently asked questions

No. Prefer’s docs say that in the current release all actions are manual: the card gives you the plan and copy-ready assets, and you or your team ship the change. If you want the work done for you, that is Prefer Managed.
Prefer lists the prioritized action engine on Growth, Scale and Enterprise; Starter does not list it. Every premium feature is switched on during the free trial, so you can try the board before you choose a plan.
Yes. In Prefer you dismiss a card with an optional reason. Prefer matches actions by signature, so a dismissed or accepted card keeps its stage when suggestions regenerate after the next tracking run.

​
Also asked as

  • What does the Action Center do in Prefer?
  • How does Prefer tell me what to fix for AI search?
  • Does Prefer give recommendations or only data?

​
Sources