My method
I learn the business model, verify the data, identify what is limiting the result and define the right order of intervention. Then I work hands-on on execution and coordinate the skills required.
Where the result is lost
Layer working as expectedCritical point to verify
Overview
The method exists to reduce decisions based on impressions. Every phase produces information or actions the next phase uses. It isn’t identical for everyone: weight and duration change with context.
Business model, constraints and shared objectives.
What data can be used and what must be fixed.
Likely causes, not just symptoms.
Order of work, between fixes and structural change.
Work done directly or coordinated, with clear ownership.
Verified effect and the decisions that follow.
Phase 01
The work starts with the company, not the ad account. It’s about what creates value and which constraints cannot be ignored.
Before opening Google Ads I want to understand what the company sells, to whom, at what margin and with which constraints. Without that, even a technically correct campaign can be optimised towards the wrong outcome.
Phase 02
Access to platforms doesn’t mean reliable data. Before using numbers to decide, I check how they are collected and what they really represent.
I don’t treat a number as reliable just because it appears in a dashboard. I check how it is collected, what it represents and whether it can be connected to the business outcome.
Margin is often the most useful indicator, but it isn’t automatically the only KPI: it depends on the business model.
Phase 03
The same symptom can have different origins. Diagnosis therefore covers the whole system, including what sits outside the ad platform.
A rising acquisition cost can come from the campaigns, but also from incomplete tracking, a weak landing page, unqualified leads or a less competitive offer.
Diagnostic map
Diagnosis starts from a stated problem and works back to the areas that could have produced it. The goal isn’t to list every hypothesis, but to rule out the ones the data doesn’t support.
Stated symptom: acquisition cost has increased
Structure, bids or queries no longer aligned with the goal.
Duplicated or missing conversions distorting the comparison.
Traffic arrives but the journey loses already qualified users.
Contacts increase but don’t turn into sales.
The proposition is less competitive than others in the auction.
Response times or handling that reduce the close rate.
Phase 04
Interventions don’t share the same cost, risk or payback time. Separating them avoids postponing obvious fixes and avoids selling a fix as a growth strategy.
I look for fast results where obvious corrections exist, but I don’t mistake a quick win for a growth strategy.
Low-risk work with an effect that can be verified quickly.
Work that requires design, time and involvement from other functions.
Shared objective: business results that hold over time, not an isolated peak.
Phase 05
Most of the work is operational. Analysis supports the decision, but the result arrives when the work is actually done.
I don’t just hand over a list of recommendations. When the project requires it, I run the implementation myself or coordinate whoever has to deliver it.
Even when several people are involved, it stays clear who does what, by when and against which objective.
Phase 06
After implementation the work continues with a regular review: what produced an effect, and what should follow.
Optimisation isn’t about making constant changes. It’s about understanding which changes produced an effect and which decisions should follow.
That’s why I don’t promise a constantly rising curve: I promise a way to understand what is happening and what is worth doing.
KPIs and impact
Before choosing what to measure, I define what “result” means for that company. Then I build the hierarchy that connects daily work to that result.
Not every company should measure the same things. The point is identifying which numbers really describe success for their model.
The economic outcome the company wants.
Margin, revenue, cost of distribution
The indicator main decisions are made on.
Qualified leads, bookings, enrolments, customer acquisition cost
They anticipate the result and allow earlier corrections.
Direct booking share, lead-to-sale rate, lifetime value
Not goals: they show where to look.
Conversion rate, offline outcomes, data quality
Collaboration
The sequence is almost always the same, but how deep each step goes depends on the starting point.
From the first call to ongoing work
Understanding the context and whether the problem is something I can help with.
Model, constraints, objectives and work already running.
Platforms, data, contacts and working tools.
Data quality checks and diagnosis of causes.
What to address, in which order and with what ownership.
Priorities are discussed with management, not imposed.
Direct execution or coordination of the required skills.
Checking the effect and defining the next steps.
The first step isn’t buying a package. It’s understanding whether there is a problem I can genuinely contribute to.
Communication
The quality of a collaboration depends on how easy it is to exchange information when it actually matters.
Problems shouldn’t wait for the next call. When needed, we can deal with them directly by email or WhatsApp.
Principles
Stated limits
Who this is for
The method contributes most when marketing investment already exists and the question is where it is losing effectiveness.
Direct answers
They start from the business model and available data, not from the platform. They define which results count, verify whether the numbers are reliable, identify the causes and set the order of work before executing it.
The competencies involvedAd accounts, tracking, analytics, website, CRM and sales data are reviewed to understand what is measured and how reliably. The output isn’t a list of errors, but a list of justified priorities.
Request a first analysisBy comparing platform data with website, CRM and sales data. If conversion signals are incomplete or don’t represent real value, the problem is rarely solved inside the account.
Google Ads and measurementA quick win is a low-risk correction with an effect you can verify quickly, such as removing wasted spend or fixing broken tracking. Structural work changes the system: funnel, measurement, integrations or new channels.
Areas of interventionBy defining the economic outcome first and then the metrics that anticipate it. Conversions are weighted by the value they generate, so platforms receive signals consistent with margin and not only with volume.
Measurement and dataGoogle Ads, analytics, tracking, automation and analysis are handled by me directly. Development, design, content and visual production go to specialists, with coordination and acceptance criteria defined by me.
Direct and coordinated workAssignments are normally monthly and continue while they produce a useful contribution. Contracts stay flexible and the collaboration can generally be ended with 30 days’ notice.
Let’s talk about your situationThe company gets senior, continuous marketing leadership without a full-time hire: setting priorities, coordinating suppliers and connecting marketing, data and revenue goals.
Roles heldTell me how your business works, which marketing activities are already running and where you see the problem. The first conversation is there to understand whether the conditions exist to work together.