Free PPC

Toolkit

The team has tried every platform lever and CPA still will not move. Usually that means the broken layer is two or three steps away from where everyone is looking, sitting in the offer, the tracking, or the landing page. This is the map I use to find which layer to fix before touching any platform setting.

The framework

Paid media performance is produced by 8 layers. Most teams spend 90% of their time at Layer 4 when the broken layer is at 1, 2, 3, or 5.

Top-down diagnostic

Ask each question in order. Stop at the first NO and fix that layer before touching anything below.

What this framework solves

Most paid media performance problems get blamed on platform settings. The real cause is usually two or three layers away from where the team is looking. PPC teams routinely spend 90 percent of their time inside the ad platform when the biggest performance levers sit outside it. This framework maps the entire system so you diagnose the right layer before you start fixing things.

The core idea

Paid media performance is produced by eight layers, and each depends on the ones above it. Layer 1 is Strategy Foundation (offer, audience, unit economics). Layer 2 is Measurement Foundation. Layer 3 is First-Party Data. Layer 4 is Platform Levers (what most people call “PPC”). Layer 5 is the Conversion Engine (landing pages, friction, trust). Layer 6 is Experimentation. Layer 7 is External Context. Layer 8 is Business Outcomes. The critical insight: Layers 1, 2, 3, and 5 sit outside the ad platform, and that is where most performance is actually won or lost. Optimizing a downstream layer while an upstream layer is broken is the single most common waste of time and money in the industry.

How to apply it

1. Start at Layer 1: is the offer viable for paid search? If multiple traffic sources convert poorly, the offer is the problem. No platform setting fixes a weak offer or unit economics that do not support the channel.
2. Move to Layer 2: is the data trustworthy? Compare Google Ads conversions against CRM or backend data. If variance exceeds 15 percent, do not optimize on platform data until the gap closes. Smart Bidding optimizing on bad data confidently pursues the wrong outcomes.
3. Check Layer 3: is first-party data being used? Value-based bidding, customer lists, offline conversion import. If none of these are active, there is a performance ceiling that platform levers alone cannot break through.
4. Then audit Layer 4: platform settings. Campaign structure, keywords, audiences, bids, budgets, ad creative. Most PPC knowledge focuses here, but it only matters once Layers 1 to 3 are sound.
5. Inspect Layer 5: the conversion engine. Load the landing page on mobile. Count the seconds. Read the headline against the ad. Try to complete the conversion. A 2x landing page conversion rate improvement equals doubling the budget at the same CPA, but for free.
6. Apply Layer 6 discipline: are you testing or guessing? Without controlled experiments and incrementality testing, you are reading correlation as causation, especially on branded search and remarketing.
7. Factor in Layer 7: external context. A sudden CPA spike with no account changes is usually competition, demand, seasonality, or a platform update, not your account deteriorating.
8. Reframe Layer 8: outcomes are revenue and profit, not ROAS and CPA. Both ROAS and CPA hide margin, scale, and LTV. Optimizing them in isolation can cut your most profitable campaigns.

When to use it

Use this when a campaign looks correctly configured but performance is still below target, when the team has already tried “all the platform levers” and nothing moves, or before scaling spend so you do not amplify a problem that lives outside the platform. It is less useful at the very beginning of a new account where Layer 1 fit is still unproven. In that case, validate the offer and unit economics first, then return to this framework once paid search has any baseline at all.