Module 1: Thinking Like an Advanced PPC Practitioner
Lesson 2 · Module 1 · Thinking Like an Advanced PPC Practitioner
Lesson 2 objective
Learn to separate what happened, why you think it happened and what you intend to do about it.
Three Different Things
Observation is not explanation. Explanation is not decision.
A common PPC mistake is to compress three separate stages into one sentence.
For example: “Performance Max is bringing poorer traffic, so we should reduce the budget.”
That statement contains an observation, an explanation and a recommendation — but only one of them may currently be supported by evidence.
Observation
What happened?
Conversion rate decreased from 5.1% to 3.9%.
Explanation
Why?
Traffic quality may have weakened after the campaign expanded into different auctions.
Decision
What now?
Investigate traffic mix and conversion measurement before changing budget.
Evidence Discipline
Say only what the data supports.
Observation → Hypothesis → Evidence → Decision
The stronger the commercial consequence of the decision, the stronger the evidence should be before you act.
Observation
CPA increased by 24% compared with the previous 30-day period.
Hypothesis
The increase may be related to lower conversion rate from mobile traffic.
Evidence
Mobile conversion rate fell 31%, while desktop remained broadly stable and mobile traffic share increased.
Decision
Investigate landing-page behaviour and tracking on mobile before changing campaign-level bids or budgets.
A plausible explanation is still only a hypothesis until the evidence discriminates it from other explanations.
Common Analytical Errors
Where experienced marketers still get caught.
Error 1
Post hoc reasoning
A change happened after an optimisation, therefore the optimisation caused the change.
Error 2
Single-cause thinking
Assuming one visible change explains performance when several drivers moved simultaneously.
Error 3
Action bias
Feeling that a declining metric requires an immediate account change even when the diagnosis is uncertain.
Scenario
A tracking change can look like a marketing problem.
Account situation
Reported conversions decline 18% week on week. CPC is stable. Click volume is stable. Search-term mix appears similar.
Three days before the decline, a new consent-management implementation went live.
- Observation: reported conversions are down.
- Hypothesis 1: underlying demand weakened.
- Hypothesis 2: traffic quality deteriorated.
- Hypothesis 3: measurement changed.
The existence of the tracking change does not prove Hypothesis 3 — but it substantially changes what you should validate before reducing bids or budgets.
Practical Challenge
Rewrite a weak diagnosis.
Choose one statement you have recently made about a PPC account, such as:
“Broad match is causing the CPA problem.”
- Rewrite the statement as a pure observation.
- List the explanation as a hypothesis rather than a fact.
- List at least two competing explanations.
- Identify what evidence would distinguish between them.
- State what action would be justified if the evidence supports the hypothesis.
Knowledge Check
Separate evidence from interpretation.
1. Conversion rate falls after broad match is introduced. What can you conclude immediately?
2. Reported conversions fall immediately after a tracking implementation. What is the strongest next step?
3. Which statement is a hypothesis rather than an observation?
4. Why is separating observation from explanation useful?
State the observation first. Treat the explanation as a hypothesis. Make the decision only after the evidence is strong enough.
This discipline becomes increasingly important as accounts become larger and more automated.