Facebook tracking pixel Skip to main content

Paid ads

Make your page match your ad.

Starts from the ad: its headline, the promise the click was sold on, and the search term or audience it targets. It then rates the page that click lands on against the ad, with evidence, on four things: a headline that echoes the ad, one action to take, real proof near that action, and the friction in the way. An existing page gets a findings table and a rewritten hero. With no page yet, it drafts the hero, call to action, proof and objection answers the ad needs.

Request access Have us build it
  • Starts from the ad
  • Leaves the ad untouched
  • One action, one metric

At a glance

Group
Paid ads
Command
/ads-landing
Uses
The ad and its landing page
Writes
A critique or ad-matched copy
Sends
Nothing. Drafts only.

01When to use it

Use it on the page behind a paid click.

  • An ad is live, or about to be, and the page it sends people to should match it.
  • People click the ad, but the page turns few of those clicks into the action, and you want to know why.
  • You are about to raise ad spend and want the page checked before more traffic reaches it.
  • You have the ad copy and the page, or just its address, and want a tight review of match, action, proof and friction.

Not for

  • A full page written from a blank sheet, every section included. Use /landing-page
  • Checking a whole ad account, or finding what to fix on the ad side. Use /ads-audit

02What it needs

It needs the ad as much as the page.

  • The ad: its headline, main text and the offer behind the click, plus the keyword or audience it targets. Paste it, save an export, or point to your offers file.
  • The page: its live address, a draft saved in the project, or an earlier landing-page deliverable. Given only an address and no page reader connected, it asks you to paste the copy or save an export.
  • Who the ad is aimed at, from your buyer file: their problem, what makes them hesitate, and the words they use.
  • Your voice files, plus real proof: recorded results, case studies, and angles that worked before.
  • Answers to any gaps: the one action and where it leads, whether this is a critique or a new draft, where the ad runs, and any claim you insist on.
  • Your rules file, read before any channel is suggested. If it is missing, it tells you and asks you before it recommends any channel or any paid spend.

03What it writes

A critique or a draft, with the metric named.

brief.md
The ad's promise and the term or audience it targets, the page being judged, the single action and its destination, where the traffic comes from, the metric, the mode, and the acceptance criteria.
final/ads-landing.md
The mode and a summary of how the page matches the ad. Then either the findings table (lever, issue, severity, evidence, fix, estimated impact on the metric) or the drafted hero, call to action, proof and objection copy. Plus a section on measuring the result.
data/baseline.json
The metric, its current figure or null when there is none, the target, the four lever scores and the proof cited. Raw ad or page exports stay in data/ too.
Ledger row
Post-click conversion rate, sourced, with a confidence of confirmed, reported, inferred or stale. With no recorded figure, the first row reads unknown rather than zero, and the target waits for your confirmation.
Daily note
One line under Activity.

04Limits

What it never does.

  • Publish the page, change your site or its content system, or edit, pause or repoint the ad.
  • Make up proof. A missing result or testimonial becomes a flagged placeholder.
  • Plan around a channel your rules forbid, or quietly swap in a similar one. It names the channel and the reason.
  • Lift a competitor's copy or headline.
  • Read a metric as zero when nothing has measured it.

What stays with you.

  • Shipping each change to the page, and anything done to the ad.
  • Confirming the target that the first ledger row flags.
  • Settling who approves when two of your files name different people.

05Sample output

Four levers, scored against the ad.

An excerpt from a critique for a made-up window cleaning company whose ad promises online booking. A real critique carries evidence on every row and opens with the headline gaps.

final/ads-landing.md Illustrative, not a client result.
Headline match High
IssueThe headline drops the ad promise
EvidenceAd: "Book a window clean online." Page: "Welcome to Example Window Care."
FixOpen with online booking, in the ad's own words
Est. impactLarge
Single action High
IssueThree asks above the fold
EvidenceBook now, Call us and Our services sit side by side
FixKeep Book now and move the other two lower
Est. impactLarge
Proof Medium
IssuePraise with nothing specific
Evidence"Loved by local homeowners," with no names or reviews
FixPut a real, named review beside the form, or a flagged placeholder
Est. impactMedium
Friction Medium
IssueA long booking form
EvidenceNine fields, including how you heard about us
FixAsk for name, address and a date only
Est. impactMedium

Rewritten hero

  • Headline: Book your window clean online.
  • Subhead: For homes in Example County. Pick a day and book your clean.
  • Button, the only one above the fold: Book my clean.

06Related skills

Often run beside it.

07Questions

Before you run it.

Will it touch my ad or my live page?

No. It reads them and writes a draft. A person ships every change to the page, and the ad is never edited, paused or pointed elsewhere.

Can it write the whole landing page?

Not the full build. With no page in place it drafts what the ad needs: a matching hero, one call to action, a short proof block, and answers to the top objections people have after the click. A page with every section is the job of /landing-page.

How is the page measured?

By one number: completed actions divided by ad clicks, or by visits from that ad. The draft says where the number is read, such as form or checkout completions or the tracking tag on the ad traffic, and the one decision it drives: keep, fix or pull the ad. With the ad spend known, it adds cost per lead.

Have us build it

We can build it for you.

Start with the free plan. A person names the first fix.

Get your free plan Or run it yourself with the toolkit.