Learn Google Ads as a system, not a menu
Google’s own Skillshop currently separates training across Search, Measurement, Video, Shopping and other product areas. Its Search curriculum includes foundations and intermediate AI-powered Search material.
That structure reflects a real change in the job.
Google Ads is no longer one discipline called “PPC.”
Search query management, Merchant Center, Performance Max, video inventory, Demand Gen, conversion architecture and automated bidding require different competencies.
The operator needs to understand how they interact.
This path is designed around seven stages:
Search intent → Conversion value → Shopping data → PMax → Demand Gen → Experimentation → Governance
The outcome is not certification.
It is the ability to operate a mixed Google Ads portfolio without treating every campaign type as a separate universe.
Prerequisite — know the interface, skip the trivia
Before starting, the learner should be able to:
- create a Search campaign;
- understand keywords and match types;
- read search terms;
- create conversion actions;
- explain tCPA and tROAS conceptually;
- navigate Merchant Center;
- read basic campaign reports.
Do not spend advanced training time memorizing menu locations.
Interfaces change.
Operator logic should survive the interface.
Stage 1 — Search: learn intent control
Competency
Understand the relationship between:
- query;
- keyword;
- match behavior;
- landing page;
- ad;
- bidding;
- negative control;
- AI Max expansion.
Google’s current AI-powered Search training and AI Max direction make this more important, not less.
The old skill was building exhaustive keyword trees.
The current skill is deciding where exploration is economically acceptable.
Exercise
Take 200 search terms and classify them by:
- business relevance;
- intent;
- brand/non-brand;
- landing-page fit;
- margin opportunity;
- negative risk.
Then build an operating policy:
- what should be blocked;
- what should become a dedicated theme;
- what can remain algorithmic discovery;
- what requires tighter landing-page control.
Checkpoint
The learner must explain why a query can be relevant and still economically undesirable.
Stage 2 — Bidding: teach the machine the right value
Competency
Understand:
- Maximize Conversions;
- tCPA;
- Maximize Conversion Value;
- tROAS;
- value rules;
- offline outcomes;
- conversion lag;
- new-customer value.
The learner should know that bidding strategy follows conversion architecture.
Scenario
A lead account has 500 monthly leads, but only 80 become qualified and 20 become customers.
Compare three possible optimization signals:
- form submission;
- qualified lead;
- closed revenue.
Ask which one should be primary.
The answer depends on signal volume, latency and value quality.
The operator must articulate the trade-off.
Failure mode
Changing tROAS because last week’s ROAS missed target.
Targets are control parameters, not emotional reactions to a dashboard.
Stage 3 — Merchant Center and Shopping: operate the catalog
Competency
Understand product data as media infrastructure.
The learner should be able to audit:
- item ID;
- GTIN;
- brand;
- title;
- description;
- product type;
- Google category;
- price;
- availability;
- images;
- custom labels;
- promotions.
Exercise
Give the learner a catalog with:
- missing GTINs;
- stale prices;
- poor titles;
- mixed product taxonomy;
- high-margin and low-margin products;
- stock constraints.
They must prioritize fixes by business value at risk, not by warning count.
Checkpoint
The learner can explain when Merchant Center native data sources are sufficient and when a feed-management layer is justified.
Stage 4 — Performance Max: define the job
Competency
Understand PMax as a multi-inventory allocation system, not a magical campaign type.
The learner should be able to define:
- campaign economic boundary;
- product segmentation;
- asset-group logic;
- search themes;
- brand controls;
- URL controls;
- creative coverage;
- relationship with Search.
Exercise
Take an ecommerce account with three product families:
- one high-margin bestseller;
- one low-margin commodity;
- one strategic new launch.
Design the PMax architecture.
Every split must have a reason.
“Because they are different categories” is not sufficient.
A valid reason could be:
- different margin;
- different budget protection;
- different launch objective;
- different inventory;
- different conversion value.
Stage 5 — Demand Gen and video: learn demand creation
Search captures expressed intent.
Demand Gen and video can operate earlier.
That changes measurement.
Competency
The learner should understand:
- creative-first inventory;
- audience signals;
- video and image asset requirements;
- conversion versus engagement objectives;
- view-through behavior;
- assisted journeys;
- incrementality.
Exercise
Design a campaign for a product with low category awareness.
Require the learner to specify:
- audience hypothesis;
- creative concept;
- measurement KPI;
- platform KPI;
- causal validation method.
If the plan judges an upper-funnel campaign only by last-click ROAS, it fails.
Stage 6 — Measurement: build the conversion feedback loop
Google Skillshop maintains a dedicated Measurement learning area because measurement is no longer a subtopic.
The operator must know:
- Enhanced Conversions;
- Data Manager;
- consent;
- offline conversions;
- transaction IDs;
- lead imports;
- GA4 relationship;
- attribution settings;
- experiments.
Exercise
Draw the conversion path from:
ad click → site → backend → CRM → Google Ads
Mark:
- event origin;
- event ID;
- value;
- latency;
- consent;
- failure points.
Failure mode
Treating Google Ads conversion count as business truth.
The operator should reconcile with source systems.
Stage 7 — Experiments: make automation earn rollout
Competency
Use experiments to answer real operating questions.
Examples:
- AI Max on versus off;
- different bidding strategy;
- different ROAS target;
- new conversion value model;
- feed-title change;
- PMax segmentation;
- landing-page expansion.
Exercise
Design an AI Max test.
The learner must define:
- hypothesis;
- treatment;
- control;
- primary business KPI;
- minimum test window;
- contamination risks;
- rollout rule.
“Run it and see if conversions go up” is not an experiment plan.
Stage 8 — Account governance
This is the stage most platform certifications underemphasize.
A professional operator needs:
- naming standards;
- change-management policy;
- Google Ads Editor workflow;
- automation register;
- negative-keyword ownership;
- budget ownership;
- asset QA;
- feed QA;
- incident process;
- audit cadence.
Scenario
A script changes budgets at 08:00.
A media buyer changes them at 09:00.
An automated recommendation applies at 10:00.
At noon nobody knows which target is intentional.
The problem is governance, not bidding.
Capstone — rebuild a real account
The learner receives:
- Search campaign export;
- search-term data;
- Merchant Center issues;
- product margins;
- PMax structure;
- conversion actions;
- creative inventory;
- twelve weeks of performance.
They must deliver:
- account diagnosis;
- conversion-value plan;
- Search control policy;
- feed priorities;
- PMax architecture;
- Demand Gen test;
- experiment roadmap;
- governance checklist;
- ninety-day migration plan.
Suggested sequence
Weeks 1–2: Search and bidding.
Weeks 3–4: Merchant Center and Shopping.
Weeks 5–6: Performance Max.
Week 7: Demand Gen/video.
Week 8: Measurement.
Week 9: Experiments.
Week 10: Governance and capstone.
Skillshop is useful for product-specific learning and certification. CXL is useful for practitioner framing.
The Radar learning path should use both as references and still require something they cannot certify for us:
Can the operator make a defensible business decision when the platform gives them multiple technically valid options?
That is the professional standard.
