A playbook in action
Find what search is costing them — then prove you fixed it
Four playbooks, one billable loop. Connect the client’s Search Console read-only, measure the window you choose, and every leaking page comes back diagnosed to one cause with the evidence quoted and the replacement titles already written. Forward the white-label report — then re-run next month and the recoveries land on the record.
New here? Connecting Claude takes about a minute →The loop, in four moves
300 tokens a run, and Claude confirms the price before it charges. Everything else is free.
Search pulse setup · free
Connect once, read-only — then map a property per client
Connect the Google account that already has Search Console access to your clients' sites — the standard agency setup. It reads performance data and nothing else: nothing is ever written to the client's Search Console, and disconnecting removes the stored credential entirely. Claude checks where the kit stands and walks you to the picker; connecting itself happens in your browser, never in chat.
“Set up Search Pulse for Acme — is Search Console connected, and does this kit have a property mapped?”
Performance data only. Nothing is written back, and disconnecting deletes the stored credential.
Connect the agency's Google account once — it covers every client site that account can already see.
Each client kit maps to one Search Console property, so a pulse always measures the right site.
Search pulse measure · 300 tokens
You pick the window — it measures against the span before it
SEO is a time span, not a snapshot. Say “since the last pulse” and it picks up where the last one ended; give dates and it uses those; say nothing and it takes ninety days. Whatever you choose is measured against the equal-length span immediately before it, so every movement is like-for-like. Claude states the 300-token price and waits for your explicit go — it never charges quietly.
“Measure the Search Pulse for Acme, since the last pulse.”
300 tokens, stated plainly, and Claude waits for your go. It is the only paid step in the whole loop.
The run hands back a link that fills in as findings land — no login, safe to forward, live for seven days.
Then the digest: findings by lane, clicks left on the table, brand versus discovery, and what's growing.
Search pulse findings · free
One cause per finding — and the rewrites already written
Claude walks the findings worst first: the page, what the numbers did over the window, and the ONE diagnosed cause with its confidence and the evidence quoted. If the results-page photo came from an earlier run, it says so with the date — results pages change hour to hour. Every click gap arrives with three replacement titles and three descriptions ready to paste, and Claude names the one it would ship and why that lever fits this SERP.
“Walk me through Acme's findings worst first — what is the one cause behind each?”
Copy-ready, each with its persuasion lever named — and Claude's pick of the strongest, with the reasoning.
The quotable answer block, FAQ structured data aimed at the People-Also-Ask questions, and content a summary can't replace.
What's already growing, so the plan protects momentum instead of only chasing losses.
Search pulse report · free
Forward the link — and detach the last page if you want
The report is white-labeled with your agency mark and the client's logo, and states the window up front. It arrives as a clickable link, live for seven days, and prints to PDF with a client-ready filename. The final page is Recommended Actions and detaches by design: leave it in and the report is your proposal, drop it and it's a findings document. Regenerate it free from any past run.
“Generate the Search Pulse client report for Acme and send me the link.”
Then the ritual: next month’s window starts where this one ended, so the recoveries land on the record — the same fixes, scored by the next pulse.
Not an SEO audit with a new label
An audit hands over a list. This commits to a cause, writes the fix, and comes back to score it.
One cause, not a checklist
Every finding gets exactly one diagnosed cause from a fixed set of seven, with a confidence score and the evidence quoted. A list of eight maybes is a way of not committing.
Rewrites, not advice
Click gaps come back with three replacement titles and three descriptions, copy-ready, each with its persuasion lever named. A twenty-minute deploy, not a content project.
The results page, photographed
The live mobile SERP is captured for every finding — competitors, People-Also-Ask, and the AI Overview when Google shows one — and stamped with the moment it was taken.
The next pulse scores the fixes
Re-run against the same site after the work lands, and every recommendation comes back as a number. That arc is the retainer; an audit stops at the score.
Four playbooks, one loop
Run them in order the first month, then just re-run the measure and the report — the connection is already made.
Search Console connected read-only, the client's property mapped, ready to measure.
The window measured, every finding diagnosed to one cause — with the live link to watch.
Every finding turned into an action — rewrites ready to paste, plays ranked by payoff.
The white-labeled SEO diagnosis report — generated and linked, ready to forward.
This is one lane. The playbook idea runs the whole agency.
Creative, intelligence, and account work — over 150 playbooks, all run by talking to Claude, all on your brand. Search Pulse is just what it looks like up close. Pair it with AI Visibility Pulse and you cover both shelves: the one Google shows, and the one the AI engines describe.