Module 1: Thinking Like an Advanced PPC Practitioner
Lesson 6 · Module 1 · Thinking Like an Advanced PPC Practitioner
Lesson 6 objective
Learn how to decide where Google Ads automation should operate freely, where it needs constraints and where human judgement should remain dominant.
The Wrong Debate
Automation versus manual control is the wrong question.
Google Ads increasingly relies on machine learning for bidding, targeting, creative combinations, campaign expansion and budget allocation.
The practitioner decision is not whether automation is good or bad. It is whether the system has the right objective, enough useful signal and appropriate boundaries.
Automation is strongest at optimisation within a well-defined objective. Humans are strongest at defining whether that objective is commercially correct.
Control Framework
Objective → Signal → Freedom → Guardrail → Review
Give automation freedom only after defining what success means.
Objective
Signal
Freedom
Guardrail
Review
Objective
What is the system actually being asked to optimise: conversions, value, profit, qualified leads or something else?
Signal
Are the conversion inputs frequent, accurate and commercially meaningful?
Freedom
How much latitude should the system have over bids, targeting, inventory and spend?
Guardrail
What limits protect margin, brand, budget, lead quality or strategic priorities?
Review
How will you determine whether automation is improving the business outcome rather than simply hitting a platform metric?
Where Automation Excels
Machine learning can process auction complexity humans cannot.
Automation is well suited to
- Evaluating large numbers of auction-time signals.
- Adjusting bids across many auctions.
- Finding patterns across conversion history.
- Optimising toward clearly defined conversion value.
- Testing combinations at a scale humans could not manage manually.
Human judgement remains critical for
- Defining commercial objectives.
- Deciding whether conversion signals are meaningful.
- Assessing incrementality and marginal economics.
- Understanding stock, margin, capacity and strategic priorities.
- Recognising when the system is optimising the wrong outcome.
Signal Quality
Automation magnifies the value — or weakness — of your inputs.
If all conversions are treated equally, Smart Bidding may optimise aggressively toward the easiest conversions rather than the most valuable customers.
If poor-quality leads are imported as successful conversions, the system can become very efficient at finding more poor-quality leads.
Better automation often starts with better measurement rather than a different bidding strategy.
Scenario
The algorithm is hitting target CPA. The business is unhappy.
Lead-generation account
- Target CPA: £65
- Actual CPA: £61
- Lead volume: +22%
- Sales team reports lower lead quality.
- CRM shows qualified-lead rate fell from 42% to 27%.
The bidding system may be performing exactly as instructed. The problem is that the optimisation signal does not sufficiently represent business value.
The advanced response is not simply to blame automation. It is to improve the objective and signal quality.
When To Intervene
Intervene because the decision system is wrong — not because you dislike losing manual control.
Weak conversion signal
Too little volume, delayed imports or low-quality conversion definitions may reduce automation quality.
Commercial mismatch
The platform is optimising a metric that does not reflect margin, customer quality or strategic value.
Unacceptable expansion
Automation may enter inventory or demand that conflicts with brand, economics or business priorities.
Structural change
Major pricing, stock, website or tracking changes can temporarily invalidate historical patterns.
Practical Challenge
Audit one automated campaign.
- Write down the exact objective the campaign is optimising toward.
- List the conversion signals feeding that objective.
- Assess whether those signals reflect real business value.
- Identify the freedom Google currently has.
- List the guardrails currently in place.
- Define what evidence would make you trust the automation more.
- Define what evidence would justify stronger intervention.
Knowledge Check
Use automation deliberately.
1. Smart Bidding is meeting target CPA, but lead quality has deteriorated materially. What is the strongest response?
2. What is the strongest role for human judgement in an automated account?
3. When should automation usually be given more freedom?
4. What is the best summary of automation at advanced PPC level?
The question is not whether to trust automation. It is whether you have given it the right objective, signals and boundaries.
Advanced practitioners design the system Google is allowed to optimise within.