Earned-media playbooks

What a play is

Sources is the diagnosis — “AI cites G2 on five of your prompts and you’re not on the page.” A play is the treatment. Rather than leaving you to figure out what to do, it gives you a typed checklist, the exact URLs to act on, an outreach draft to start from, and — the part that makes it a loop and not a to-do — a verification window that re-measures whether the citation actually changed.

This is the off-site half of AEO that the rest of OptimizeCamp structurally can’t reach — see Impact Cards for the on-site equivalent that proves an edit or a new article worked.

The six play types

A play’s type is chosen for you from the source: if your brand is already present on the page, you get a defend play; otherwise the type follows the source type. Each type comes with its own checklist and a fixed effort estimate.

The Plays worklist

The Plays screen (/[org]/projects/[project]/visibility/plays) is the triage view — every play across the project in one sortable list. Four tiles summarize it:

(Plus an All plays tile.) A row of filter chips — Open, Suggested, In progress, Verifying, Won, All — narrows the list, and an Assignee filter appears once your workspace has more than one member. The table shows each play’s name (“{play} — {domain}”), status, step progress (e.g. 2/5), Opportunity score, effort, assignee and last-change date — sorted by opportunity, highest first.

Every row deep-links into the playbook drawer, which opens inside the Sources table (…/visibility/sources?play={playId}). The worklist is for triage and routing; the drawer is where you actually work a play.

Inside the playbook drawer

The drawer is the single working surface for one play. Top to bottom, it holds:

• Why now — the evidence the play was built from: “Cited N× across M of your prompts,” the prompt chips, any competitors in those answers, and a presence pill with a Check presence button if it’s unknown.
• Targets — the actual cited URLs to act on, as links (Open thread / Open listing / Open roundup / Open profile, depending on the play). If none are captured yet, “the next tracking run fills these in.”
• Outreach draft — a Draft outreach copy button that writes a starting subject and body for you (more on this below).
• Steps — a checkable list (Steps · {done} of {total}); ticking a box is a real state change and, on a suggested play, starts it.
• Outreach log — a freeform note (“who you contacted, when, what they said”) that survives everything, even a reopen.
• Assignee — hand the play to a teammate (shown once the workspace has more than one member).
• Status controls — the buttons that move the play through its lifecycle, described next.

The lifecycle

A play moves through a fixed set of states, each shown as a status chip:

Suggested In progress Done Verifying Won No change Dismissed

From any open state you can Dismiss a play with a reason (“Not relevant to us,” “Not worth the effort,” “Already tried it,” “Other”). A closed play can be reopened — Reopen playbook for a dismissed one, Run it again for a won or no-change one — which resets the verification window and drops it back to in progress.

How plays are ranked — the Opportunity score

Every play carries a 0–100 Opportunity score — the same score the Action Center ranks by — so the highest-leverage sources float to the top of the worklist. It blends four things: how often the domain is cited, how many of your prompts it’s cited on, whether you’re absent from the page (a confirmed absence weighs most, an unknown counts for half), and whether competitors show up in those answers. Citation-heavy, many-prompt, you’re-absent, competitors-present is the maximum- opportunity shape.

The verification window & Earned Impact Cards

When you mark a play done, OptimizeCamp watches the next tracking runs on the affected prompts and, once the window closes, writes an Earned Impact Card — the before/after receipt for the play. It reports whether the domain cited your brand before vs. after, whether your brand appeared on the page before vs. after, any new prompt coverage, and how many runs it measured over. The verdict is won if the domain now cites your brand, your presence flipped from absent to present, or you picked up new prompt coverage; otherwise it’s a measured no change (a domain that stopped citing you is flagged as a decline inside the card).

Outreach drafts

Open plays offer a Draft outreach copy button that generates a subject and body to start from — grounded in your brand profile and the actual cited page. Anything you must supply yourself is left as a {placeholder}, and the copy carries the same disclosure and anti-astroturfing guardrails as the steps.

Without a live model key you still get an honest template draft (free, unlimited). A real AI draft is metered: it counts against your plan’s monthly AI-draft allowance — Starter 5, Pro 20, Agency 50. Template drafts and failed calls never count. At the cap the drawer tells you so and points you to your next billing period or an upgrade — see Plans, quotas & credits.

Where plays surface elsewhere

You don’t have to live on the Plays screen. The Action Center places plays in the appropriate lane: a “Start playbook” opportunity when AI cites a domain on several prompts, a “Resume playbook” card for one you’ve part- finished, and a “See the impact” card once a play resolves. Won plays also land in the Wins feed (“Earned a citation on {domain} after your {play}”). A play that’s verifying shows no card — the loop owns it until it has a verdict.