Ninja Scripts Academy
Google Ads from top to bottom: basic, intermediate and advanced. With real examples, best practices and the gaps the scripts fill. 🔄 Syllabus reviewed and updated every month with the latest Google Ads news · 🎧 Every lesson also in audio · 🖍️ Highlight and save your favourite lessons to come back to them.
🌐 Lessons are in Spanish for now; the English version is in preparation.
Ninja Scripts Suite subscribers get 100 % of the three levels, every new lesson each month and the study tools: 🎧 audio for every lesson, ⭐ favourites and 🖍️ highlights. See the subscription →
Intermediate training
60 lessons · 9 h 33 minMatch types in 2026: what really triggers your ads 🔑 With a free account
Exact, phrase and broad no longer mean what the 2018 manual said — how intent-based matching works today and how to keep control.
Search term mining: from thousands of rows to ten decisions 🔑 With a free account
The method for analysing terms at scale — export, group by n-grams and roots, classify by intent, cross-reference with cost and conversions — and the worksheet that turns an unmanageable report into a short list of actions.
Negativisation scenarios: rules that decide for you (and how to calibrate them) 🔒 Subscribers
How to turn judgement into rules — cost with no conversions, low CTR with impressions, runaway CPA, semantic irrelevance — with thresholds tied to the target CPA, rotating shared lists and the safeguards that stop you negativising what sells.
The white list: controlling traffic the other way round 🔒 Subscribers
The inverse approach to negatives — define what you DO want and block everything else —, when it is the right choice, how it is built and maintained, its risks and how to combine it with scenarios.
Broad match with Smart Bidding: how to test it without burning money 🔒 Subscribers
Why modern broad only works with automated bidding, how to set up a clean experiment (50/50, one ad group, negative seatbelt), what to measure, when it wins and when it does not, and the account pattern that integrates it.
Dynamic Search Ads (DSA): the demand finder that reads your website 🔑 With a free account
How DSA work (Google generates the headline and the destination from your pages), their three profitable uses — discovering terms, covering large catalogues, filling gaps —, how to control them with negative targets and how to turn what they discover into keywords.
Smart Bidding from the inside: signals, prediction and what the algorithm cannot know 🔑 With a free account
How Google turns the signals of each auction into a bid, what data feeds it, the «top signals» in the bid strategy report, and the algorithm's three blind spots that are still your job.
Portfolio strategies and shared budgets: pooling data without losing control 🔑 With a free account
When to group campaigns into a portfolio (volume, a common target), the maximum and minimum bid limits, how to combine a portfolio with a shared budget, the typical mistakes and a portfolio structure for a mid-sized account.
Seasonality adjustments and data exclusions: telling the algorithm what you know 🔒 Subscribers
The two tools for correcting Smart Bidding from the outside — announcing a short spike or drop in the conversion rate, and wiping a period of broken data from its memory — when to use them, when not to, and how to set them up without doing damage.
From manual bidding to Smart Bidding without sinking the account: the transition plan 🔒 Subscribers
How to move from manual or enhanced CPC to target CPA/ROAS in four phases — prepare the measurement, choose the bridge strategy, set the first target from the data, and the four weeks of watching — with the normal symptoms and the alarming ones.
The judge: judging each campaign by its business target, not by the strategy target 🔒 Subscribers
The difference between what Smart Bidding chases and what the business needs, the «cross-mode» case (a campaign on target CPA judged on ROAS and vice versa), reference conversions, and the method for deciding when to move targets and budgets — and by how much.
Diagnosing Smart Bidding: symptoms, causes and the bid strategy report 🔑 With a free account
The symptom table (spend that collapses, CPA that climbs, endless learning, volume that vanishes) with its likely causes and how to check each one, how to read the bid strategy report, and the checklist to run before blaming the algorithm.
The three versions of Quality Score: the warning light, the doorman and the one that charges you 🔑 With a free account
Why the 1-to-10 number is a simplification that is never used in the auction, what the ranking QS and the CPC QS are that really do decide position and price in every search, and why «Quality Score no longer matters» is a sentence that depends on which of the three you mean.
The three subfactors do not weigh the same — and relevance, to Google, is called CTR 🔑 With a free account
How much each Quality Score component weighs according to the most quoted reverse engineering in the industry, why relevance is a means and CTR is the end, the four historical CTRs Google combines without showing you, and the tactic of «paying more at the start to pay less later».
The small print of Quality Score: one QS per term, per device, per area — and what does NOT affect it 🔒 Subscribers
Every search term carries its own hidden QS (and how to «pull the good ones out of the bag»), the separate QS figures by device, network, geography and time of day, vertical and horizontal relevance, the myths about what affects it, and why the landing page can only hurt you.
Where QS hits: the position leapfrog and how much money is at stake 🔑 With a free account
The threshold mechanism that lets an advertiser with a QS of 9 and the lowest bid leapfrog someone bidding twice as much, the non-linear relationship between QS and CPC, and a calculation of how much more (or less) the same budget buys depending on the account's effective QS.
The relevance chain: the golden rules and the recipe for lifting a keyword 🔒 Subscribers
The six rules distilled from twenty years of accounts (with CTR as the goal of all of them), the ad test that looks at CTR × conversion and not CTR alone, and the step-by-step procedure for rescuing a specific keyword — with its small print: it works, it takes months and that is why almost nobody does it by hand.
The 30-day plan: getting Quality Score out of the drawer with automated discipline 🔒 Subscribers
Four weeks to turn QS into a routine — the X-ray, the bleeding, the structure and the components — what to look at each week, what to decide and what to let the machine do, and the Monday ritual that keeps the improvement going afterwards.
The hinge of the account: the ad is the only thing that touches all five metrics at once 🔑 With a free account
The full chain — copy → CTR → Quality Score → CPC → impressions → conversion — and the part almost nobody counts: the ad does not just attract clicks, it decides which clicks it attracts. Why optimising bids with mediocre ads is tuning the engine with the handbrake on.
«They already work, don't touch them»: the two errors behind the most expensive belief 🔑 With a free account
Why «it works» does not mean «it is the best possible», the five forces that degrade an ad even if you never change a comma (competition, wear-out, language, searches, seasonality), the reigning champion that defends its title every month, and how the RSA changed the unit of testing.
What a manager cannot do (even if they want to): the five walls of testing by hand 🔒 Subscribers
The history Google does not give you and nobody keeps, the mistakes repeated for lack of memory, the paraphrases that pass for tests, the ad judged as a whole instead of piece by piece, and the impossible discipline — plus the archive sheet that solves the first four.
The continuous optimisation cycle: the scientific method applied to ads 🔒 Subscribers
Hypothesis, exposure, measurement, verdict and memory — the cycle that never ends — with the three rules that make it work, the calendar for running it by hand on a 20-group account, and why it pays like compound interest (in both directions).
The map of angles: how to generate hypotheses that teach you something 🔑 With a free account
The ten message angles and how they are tracked per ad group (tried / pending / won / lost), where good hypotheses come from — search terms, landing page, history, competitors — and the template for writing a challenger that is not a paraphrase.
Automated discipline: how you work with the RSA Optimizer in practice 🔒 Subscribers
The archive the script keeps for you, the AI proposals that learn from the ad group's history, the connection with real Quality Score, the «nothing goes live without your checkbox» rule, the 15-minute weekly flow and the safeguards that stop testing turning into noise.
Shopping from the inside: how Google decides which product to show (with no keywords) 🔑 With a free account
The search ↔ product matching built from the feed, the factors that decide which listing shows and in what position (title, relevance, price, bid, history), the Shopping auction versus the text one, and what you really control.
Shopping campaign structure: product groups, priorities and the architecture by intent 🔑 With a free account
How to subdivide the catalogue into product groups (category, brand, custom labels), what campaign priority does and how to use it with negatives to separate brand, generic and specific-product searches — the classic three-tier architecture.
Search terms and negatives in Shopping: your only control over searches 🔒 Subscribers
How to read the search terms report of a Shopping campaign (with no keywords to explain it), the patterns that give away a feed with badly chosen vocabulary, negatives by priority tier, and the universal negative list of an ecommerce business.
Bidding in Shopping: target ROAS by margin, real sales and POAS 🔒 Subscribers
Why a single ROAS for the whole catalogue is a mistake, how to set targets by margin group, the trap of judging Shopping by revenue (real sales, returns, cancelled orders) and the indicator that should be in charge: return on profit.
Merchant Center health: disapprovals, mismatched prices and the suspension that arrives without warning 🔑 With a free account
Where to see the state of the catalogue, the most frequent disapproval reasons and their fixes, automatic item updates, the policies that suspend whole accounts and the weekly routine that stops you losing Shopping on a Monday morning.
Product-level management: acting on the catalogue with real sales data 🔒 Subscribers
The Shopping workflow at scale — per-product analysis with real sales and margin, the profitability traffic light, the three checkbox actions (exclude, negativise, improve titles), TEST mode and verification — and how the work is split between the script and you.
Anatomy of the feed: required attributes, recommended ones and the ones that actually move sales 🔑 With a free account
The product data specification explained by usefulness (not in alphabetical order), the ways to submit the feed (file, platform, API), the frequency each type of shop needs and the structural mistakes that limit visibility without disapproving anything.
Product titles: the formula by vertical and the 70 characters people actually see 🔒 Subscribers
How to write the title Google matches and the user reads — the order of elements by type of product, what goes in the first 70 characters, what must not be there (capitals, promos, adjectives) — and how to use the description for the vocabulary that does not fit.
Identifiers and categories: GTIN, brand, Google category, product type and variants 🔑 With a free account
What each identifier does for your visibility (and what you lose with no GTIN), when to declare that a product has no code, how to choose the Google category and build your own product type, and how to group variants so Google shows the right size and colour.
Feed rules and supplemental feeds: transforming the catalogue without touching the website 🔒 Subscribers
How to modify, complete and enrich the feed from Merchant Center — attribute rules (conditions and transformations), supplemental feeds (a spreadsheet that overwrites columns by ID), calculated custom labels — and the workflow that lets marketing change titles and margins without waiting for a developer.
Images, prices, availability and shipping: the attributes that decide the click and prevent disapproval 🔑 With a free account
Which image wins the listing (and which one Google disapproves), how to send price and sale price with no discrepancies, availability and stock in near real time, shipping and returns settings, and the website structured data that keeps everything in sync.
Continuous feed optimisation: testing titles and attributes the way you test ads 🔒 Subscribers
The feed is not configured once — it is cultivated. How to measure performance by attribute (which titles, categories and labels sell), the title test on a pilot category, the monthly feed routine and the feed ↔ terms ↔ sales circuit that closes the module.
Google's audience map: your data, in-market, affinity, custom and demographics 🔑 With a free account
Every audience type Google offers, which signal it comes from, where you can use it (Search, Display, video, Demand Gen, PMax) and the decisive difference between «Observation» and «Targeting» — the one that separates learning from restricting.
Remarketing in depth: behaviour-based lists, membership duration, exclusions and frequency 🔒 Subscribers
How to build lists that mean something (not «all visitors»), membership duration matched to the buying cycle, excluding people who already bought, capping frequency so you remind rather than stalk, dynamic remarketing and customer lists — and how to judge whether remarketing adds anything or just claims credit for sales.
Optimized targeting and audience signals: when Google widens your audience on its own 🔑 With a free account
What Display's «optimized targeting» does (it is on by default), how it differs from the audience signals in PMax and Demand Gen, when to leave it on and when to switch it off, and how to measure whether the expansion brings customers or just clicks.
Placements and exclusions: where Display money goes and how to shut the door 🔒 Subscribers
How to read the «Where ads showed» report, the signs of a junk placement (farms, filler apps, non-human traffic), content and category exclusions, account-level placement exclusion lists (the only ones PMax obeys) and the routine that keeps them up to date.
Display creative: responsive ads, one message per audience and creative fatigue 🔑 With a free account
How to build a responsive display ad that works (images, logos, headlines, descriptions), why every audience deserves its own message, how to read the asset report, spot creative fatigue and rotate before your CTR collapses.
Measuring Display honestly: view-through, windows, incrementality and experiments 🔒 Subscribers
Why view-through conversions inflate the numbers, which windows to use, the right question (what would have happened without the campaign?), the three methods for answering it — holdout experiment, geo experiment and reading the account total — and how to present Display to whoever pays for it.
Offline conversions: telling Google what happens after the form 🔑 With a free account
Auto-tagging and the GCLID, how to store the click identifier in your CRM, how to upload the conversions that happen off the website (qualified lead, sale, amount) by spreadsheet, API or integration, enhanced conversions for leads, and the deadlines and limits you have to respect.
Conversion value strategy: what each conversion is worth (and how to make Google chase it) 🔒 Subscribers
How to assign values that reflect real money — fixed by lead type, dynamic by order, margin-based, by probability of closing — the 50% rule for navigation conversions, the move to value-based bidding in lead generation, and the mistakes that make the value lie.
GA4 and Google Ads: linking, importing conversions, sharing audiences — and why they never match 🔑 With a free account
What the link gives you (GA4 events as conversions, GA4 audiences in Ads, Ads data in GA4), when to measure with GA4 and when with the Ads tag, the attribution and window differences that explain the discrepancies, and the healthy setup that stops you counting twice.
Advanced attribution: what data-driven attribution really measures and which decisions it supports 🔒 Subscribers
How data-driven attribution works, what changes between last click and DDA in the reports (and what doesn't), conversion paths and windows, the limits of any attribution inside Google Ads, and the rule for deciding budgets: attribution to allocate, incrementality to confirm.
Lead quality: scoring every contact and teaching Google the difference between a customer and a browser 🔒 Subscribers
Why the form is not the conversion, how a lead score is built from three layers (behaviour on the site, positive actions, source campaign), how it is recalibrated with the CRM, and how it is sent back to Google as a value so Smart Bidding optimises towards customers.
The advanced measurement audit: the 25-point list and the map of what reaches Google 🔑 With a free account
The full audit of a mature account — tag, consent, enhanced conversions, offline, values, duplicates, GA4, attribution — organised around a map of «which signal reaches Google, from where and with what value», plus an introduction to server-side tagging.
Anatomy of impression share: how much of the market you buy and what holds you back 🔑 With a free account
The five share metrics (search IS, top, absolute top, lost IS to budget, lost IS to rank), how they are calculated and how to read them together to know whether you can grow, at what cost and in which direction — the foundation of all planning.
Response curves: diminishing returns and the marginal cost of the next conversion 🔒 Subscribers
Why doubling the budget does not double conversions, how to draw the spend → conversions curve of a campaign from your own data and with the Performance Planner, what marginal CPA is and how it decides where the next euro goes — the concept that separates managing from spending.
The Performance Planner: useful forecasts (and their limits) 🔑 With a free account
What the Performance Planner does, how to build a monthly plan and read its curves, which campaign types it supports and which it no longer does (2026 changes), how to use it to negotiate budget and why its figures are a starting point, not a promise.
Splitting the budget across campaigns: the marginal method, the rules and the calendar 🔒 Subscribers
From a monthly total to daily budgets per campaign — the six-step method based on target attainment, lost IS and marginal CPA, the ring-fenced budgets (brand, remarketing, testing), how to handle seasonality and the mistakes that throw the split off.
Controlling monthly spend: pacing, variances, alerts and a month-end with no surprises 🔑 With a free account
How to track spend against plan day by day (expected pace, running total, projection), which variances are normal and which are warning you about a problem, the alerts worth having, what to do mid-month and how to close without overshooting or undershooting.
Annual planning: seasonality, peaks, reserve and a budget that grows with results 🔒 Subscribers
How to build the twelve-month plan from your history (a seasonality index per month and per campaign), when to raise ahead of a peak and by how much, the reserve for the unexpected and for testing, budget tied to goals (not to a fixed percentage) and how to present it to whoever signs it off.
What a Google Ads script is: a program that lives inside your account 🔑 With a free account
What scripts are (JavaScript run by Google inside the account), what they can and cannot do, how they are authorised and scheduled, the time and quota limits, the difference between account and manager account scripts, and why they are the most powerful automation tool any advertiser can reach without programming.
What the scripts solve: the problem → script map of the whole course 🔑 With a free account
The gaps in Google Ads seen in the earlier modules, sorted by problem, with the Suite script that covers each one, what it does every night, what you decide and what you see in your spreadsheet — the practical index for working out which to install first for your account.
Installing a Suite script step by step: from the panel to the first run 🔒 Subscribers
The two installation models (self-install from the panel, or installed by us in the MCC), the blank spreadsheet and who to share it with, pasting the loader into Google Ads, authorising, the preview, the scheduling and checking the first run — with the typical mistakes at each step.
Anatomy of the loader, the licence and the security: why the code never circulates 🔒 Subscribers
The three pieces of a Suite script (bootstrap, loader, server-side code), how the loader downloads and runs the code on every execution, what the licence checks and why it fails open, which data leaves your account and which does not, and how to audit for yourself what a script does.
Reading the scripts' spreadsheets: dashboards, traffic lights, checkboxes, history and the manual 🔑 With a free account
The shared anatomy of any Suite script's spreadsheet — Config, a Dashboard with a traffic light, action tabs with checkboxes, history by ID, Diagnostics and the Manual — how to read it in five minutes, what each colour means and how to approve an action without fear.
Automating with judgement: rules, scripts and AI — what to delegate, what to supervise and what never to automate 🔒 Subscribers
The three levels of automation (automated rules, scripts, generative AI) and what each is for, the four design rules of a safe automation, how to introduce one without breaking anything (shadow, test, checkbox, gradual rollout), what not to automate and how to measure whether the automation is paying off.