PPC Supreme
Free Claude skills · Google Ads

Twenty Claude skills I built for Google Ads. Yours, free.

These are the skills I use to interrogate real accounts: where the next pound should go, whether smart bidding is drifting, which campaigns actually make money once the CRM has its say. Each one is tested, read only, and works on any account in any industry. Download, install, attach an export, ask in plain words.

Download the full run guideDownload all 20 skills

2-page essentials guide covering setup, the key rules and all twenty skills. The zip contains every skill, organised by pack.

Read only, always

Every skill analyses and recommends. Nothing changes budgets, bids or campaigns. Each recommendation names an owner, the upside and downside, and how to roll it back.

Evidence before opinion

Every finding is tagged Observed, Inferred or Modelled, every estimate carries a low, central and high range, and weak data means the skill says so instead of guessing.

Install in seconds

Save each .skill file in chat or upload it in the Skills area of settings. It then works in every new chat, in any industry, with your currency and stage names.

Advanced growth and bidding intelligence

What would happen if you changed the bidding, before you touch the live account.

Shadow Bidding Engine

Estimates what an alternative target, CPC limit or budget would probably have delivered, measured on qualified leads and sales, without touching the live account. Ends with an experiment to validate the best alternative.

What to upload
90 days of daily campaign performance, bidding targets and change history, CRM outcomes, and the target or budget alternatives you want to compare.
What you get back
Modelled volume and cost ranges for each alternative, a comparison with current bidding, and a validation experiment.

Example prompt

“What would happen if we moved tCPA from £X to £Y on our two main campaigns?”

Download skill

Marginal Auction Opportunity Engine

Finds where the next block of spend is most likely to buy valuable extra business, separating budget-limited headroom from rank-limited campaigns, and proposes moves within a ceiling.

What to upload
90 days of campaign performance with impression share lost to budget and rank, campaign values from your CRM, and your monthly budget ceiling.
What you get back
Ranked spending opportunities, estimated extra spend each campaign could absorb, and likely commercial returns.

Example prompt

“Where should the next £5,000 a month go?”

Download skill

Smart Bidding Behaviour Fingerprint

Learns how each bid strategy responds to target, budget and goal changes, how long it takes to settle, and flags drift such as rising CPC without better downstream quality.

What to upload
Six months of daily bid strategy performance, target and budget change history, CRM outcomes and conversion delays.
What you get back
How each strategy responds to changes, its usual settling time, and signs of rising costs without better lead quality.

Example prompt

“Fingerprint our bid strategies. How big a target change can they absorb?”

Download skill

Intraday Demand and Budget Survival Model

Predicts whether campaigns run out of budget before the day's most valuable hours, using hourly spend, downstream value by arrival hour and team capacity.

What to upload
Eight weeks of hourly campaign performance, daily budgets, CRM outcomes by enquiry arrival hour, and team capacity.
What you get back
A budget survival verdict, cumulative spend curve, and recommendations to protect your most valuable hours.

Example prompt

“Are we running out of budget before our best hours?”

Download skill

Change Blast-Radius Simulator

Before changing a shared strategy, conversion goal, budget, negative list or asset, maps every campaign it reaches, rates the risk and writes the staging and rollback plan.

What to upload
Your proposed change, campaign list, shared budgets and bid strategies, shared negative lists, and recent performance.
What you get back
A map of affected campaigns, risk ratings, and a staged change plan with a rollback procedure.

Example prompt

“We want to switch the account default goal from form fills to qualified leads. What does that touch?”

Download skill
Hidden growth opportunities

The demand, drift and competitor pressure your account reports never show you.

Hidden Demand Discovery Engine

Finds small, growing clusters of searches the account does not yet cover, judges likely quality, names the campaign that should own each theme and whether it needs its own landing page.

What to upload
Six months of search terms and campaign performance, Search Term Insights, Search Console queries, and current keywords and negatives.
What you get back
Emerging search themes, evidence of likely quality, and suggested campaign ownership and landing pages.

Example prompt

“Find new search themes we are missing.”

Download skill

Search Intent Migration Detector

Watches what each established keyword actually matches over time, and flags keywords drifting towards research or irrelevant searches before quality falls further.

What to upload
Six months of monthly search terms by keyword, match types and costs, plus CRM outcomes by keyword or ad group.
What you get back
Keywords whose search intent is drifting, a cost weighted query mix analysis, and suggested negatives.

Example prompt

“Our top keywords are bringing weaker leads but nothing has changed in the account.”

Download skill

Competitor Pressure Attribution Engine

Weighs four explanations for rising CPCs (competition, matching, relevance, demand) plus your own changes, and ranks them with supporting and contradicting evidence.

What to upload
12 weeks of Auction Insights, weekly keyword and campaign performance, search terms, change history, and demand trends.
What you get back
Ranked explanations for rising CPCs, supporting and contradicting evidence, and a test to check the leading explanation.

Example prompt

“CPCs are up 25% in six weeks. What is driving it?”

Download skill

AI Max Expansion Quality Monitor

Separates traffic reached through AI Max or expanded matching from established keyword traffic, and judges whether the new traffic turns into qualified leads and sales.

What to upload
Search terms with match source, campaign performance, Search Term Insights, landing page reports, and CRM outcomes.
What you get back
A split of established and expanded traffic, a commercial quality verdict, and suggested exclusions for irrelevant themes.

Example prompt

“We switched on AI Max six weeks ago. Is the new traffic any good?”

Download skill

Campaign Structure Evolution Engine

Tests whether campaigns should be merged or split, based on data density per bid strategy, distinct intent and different downstream outcomes, and designs the test before any restructure.

What to upload
Your campaign and ad group structure, bidding settings, 90 days of performance, CRM outcomes, and query overlap reports.
What you get back
An evidence based merge or split recommendation and a controlled experiment plan before any restructure.

Example prompt

“We have 14 search campaigns and several feel data-starved. Merge or split?”

Download skill
Commercial intelligence

Connecting ad spend to what actually happens in the CRM afterwards.

Sales Capacity-Aware Acquisition Engine

Connects ad volume to how fast the team can handle leads, estimates the value lost when enquiries outrun capacity, and says whether to grow ads, shift hours or fix capacity first.

What to upload
8 to 12 weeks of CRM leads with arrival and response times, sales progression, and your team’s handling capacity.
What you get back
A capacity verdict, estimated value lost through slow responses, and whether to grow spend, shift hours or fix capacity first.

Example prompt

“Should we scale while the sales team is stretched?”

Download skill

Conversion Maturity and Revenue Nowcaster

Estimates how recent leads will turn out using how earlier cohorts matured, so new campaigns are not judged too early. Expected and actual results are always shown separately.

What to upload
At least 12 months of CRM leads with stage dates and values, weekly campaign spend, and recent sales process changes.
What you get back
Expected sales and CPA ranges for recent leads, with actual matured results kept separate from estimates.

Example prompt

“How will last month's leads turn out?”

Download skill

Lead Value Distortion Detector

Finds campaigns that look strong on cost per lead but produce leads worth little, by ranking campaigns four ways: CPL, cost per qualified lead, cost per sale and contribution.

What to upload
CRM leads with stage dates and sale values, campaign spend, and lead maturation curves or nowcaster outputs.
What you get back
Campaign rankings by CPL, cost per qualified lead, cost per sale and contribution, exposing cheap leads with poor value.

Example prompt

“Which campaigns really make money?”

Download skill

Duplicate Demand and Lead Recycling Engine

Separates genuinely new prospects from repeat enquiries, returning prospects and existing customers, and shows where conversions are counted twice.

What to upload
CRM leads with a consistent contact ID, enquiry dates, sources and campaigns, plus existing customer records.
What you get back
The share of genuinely new prospects, repeat enquiries and existing customers, and where conversions are counted twice.

Example prompt

“How many of our leads are genuinely new people?”

Download skill

Profit-Weighted Conversion Feedback Engine

Designs conversion values from each lead's expected contribution, using only signals known at lead time, calibrates them, and plans how to feed them back to Google Ads.

What to upload
12 months of lead history with signals available at enquiry time, outcomes and sale values, plus your conversion setup.
What you get back
Proposed lead values based on expected contribution, calibration against actual outcomes, and a Google Ads feedback plan.

Example prompt

“Design conversion values so bidding chases better leads.”

Download skill
Experimental and autonomous intelligence

Deciding what to test, what caused a shift, and what to do this week.

Experiment Opportunity Prioritiser

Scores candidate tests by expected value, cost, statistical power, risk and time to answer, so the next test answers the most valuable question.

What to upload
Your candidate tests and hypotheses, baseline campaign volumes, CRM progression rates, and lead or sale values.
What you get back
A ranked test shortlist weighing business value, statistical power, cost, risk and time to get an answer.

Example prompt

“Which of these should we test first?”

Download skill

Causal Change Investigator

Works out what most likely caused a performance shift, comparing affected and unaffected campaigns, checking seasonality, tracking and change history, and names the test that would confirm it.

What to upload
Daily results for affected and unaffected campaigns, change history, seasonality, Auction Insights, and CRM outcomes.
What you get back
Ranked possible causes with evidence for and against each, plus a test to confirm the most likely explanation.

Example prompt

“Qualified leads fell 30% from 29 July. What caused it?”

Download skill

Creative Persuasion Intelligence

Classifies ad messages into persuasion themes such as speed, certainty, price, convenience and trust, relates them to downstream quality, and produces a ranked creative test plan.

What to upload
Asset reports with performance labels, 90 days of ad results, CRM outcomes, and any brand guidelines.
What you get back
Ad messages grouped by persuasion theme, their relationship to sales quality, and a ranked creative test plan.

Example prompt

“Which ad messages bring customers who buy?”

Download skill

PPC Decision Memory

Keeps a log of significant decisions, what was expected and what happened, and checks new proposals against precedents. Works in three modes: log, review and precedent check.

What to upload
Your decision log in CSV or Markdown, with previous changes, expected results and actual outcomes.
What you get back
An updated decision record, a review of expectations versus results, or a precedent check for your next proposal.

Example prompt

“Log this decision and check it against what has worked before.”

Download skill

Autonomous Growth Strategist

Combines the other skills' outputs into one weekly decision: the binding constraint, two to four options with upside and downside ranges, one chosen move, and a plan that needs your approval.

What to upload
Your latest tracking verdict, recent skill outputs, weekly performance, CRM outcomes, decision log and business constraints.
What you get back
The main growth constraint, two to four options with upside and downside ranges, and one recommended move for your approval.

Example prompt

“Run the weekly strategy read across these outputs.”

Download skill

How I run them

Weekly

Autonomous Growth Strategist, after a tracking readout. If tracking is broken, the week's move is fixing measurement.

Monthly

Marginal Auction Opportunity Engine, Conversion Maturity and Revenue Nowcaster, Lead Value Distortion Detector, Smart Bidding Behaviour Fingerprint.

Before a change

Change Blast-Radius Simulator, Shadow Bidding Engine, Experiment Opportunity Prioritiser, and a PPC Decision Memory precedent check.

When something moves

Causal Change Investigator, Competitor Pressure Attribution Engine, Search Intent Migration Detector.

Rather I ran them on your account?

These skills are the same thinking I bring to a full audit: bidding, structure, tracking and what your leads are actually worth. If you want the answers without doing the running, start here.

Request a free account audit