96 tools and a book to actually run AI in your Google Ads accounts.

Inheriting an account is the highest-leverage moment in any client engagement. Build trust in week one and you have 12 months of runway. Move too fast on two weeks of data, or sit too long on real signal, and you spend 90 days catching up to your own onboarding. Here is the cadence that keeps the action intensity matched to the data.

The framework

Phase transitions are driven by data milestones, not calendar dates. Each phase has its own posture.

Five Foundations (verify all before any optimization)

1. Conversion tracking firing correctly (conversion-tracking-audit, measurement-playbook for 3-way validation)
2. Location targeting set to “Presence” not “Presence or Interest”
3. Attribution model appropriate to sales cycle
4. Budget distributed by priority
5. Brand terms identified and protected with a dedicated campaign. Snapshot the account state with ppc-account-health-check

Bidding strategy by phase and volume

Inherited account differences

• Observe 1 to 2 weeks before sweeping changes
• Document everything (screenshots, exports)
• Treat historical data from a broken setup as unreliable
• No weekend rebuilds

What this framework solves

The first 90 days of a Google Ads engagement decide whether the client stays or leaves. Aggressive optimization on two weeks of data is malpractice. Conservative caution at Week 10 with 200 conversions of evidence is cowardice. Most onboarding failures are mismatch between data maturity and action intensity, in either direction. This framework defines what is appropriate at each phase so you stop scaling on noise and stop sitting on signal.

The core idea

Data confidence determines what you can do, and this is not negotiable. Match action intensity to data maturity: conservative during stabilization (weeks 1 to 4, when data is thin), balanced during optimization (weeks 5 to 8, when patterns emerge), aggressive during growth (weeks 9 to 12, when evidence supports confident decisions). Phase transitions are driven by data milestones, not calendar dates: stabilization exits at 30+ accumulated conversions, not at week 4. New accounts and inherited accounts also need fundamentally different treatment. New means conservative bids and tight match types until data exists. Inherited means a forensic baseline audit, then sequential fixes, never a sweeping weekend rebuild.

How to apply it

1. Verify the Five Foundations before any optimization. Conversion tracking firing correctly. Location targeting set to “Presence” not “Presence or Interest.” Attribution model appropriate. Budget distributed by priority. Brand terms identified and protected.
2. Capture the baseline snapshot in week one. Every meaningful metric, frozen in time. Without it you cannot prove improvement, and without proven improvement you cannot retain the account.
3. Run stabilization (weeks 1 to 4) for waste prevention, not optimization. Search term review at least weekly with aggressive negatives for alignment and audience waste. No keyword pause decisions yet, except extreme cases. No smart bidding yet on low-conversion campaigns.
4. Match bidding strategy to volume. Below 15 conversions/month: Manual CPC or Maximize Clicks. 15 to 30: Enhanced CPC. 30+: Target CPA or Maximize Conversions becomes eligible. 50+ with values: Target ROAS. Premature smart bidding produces erratic performance that looks like a bidding problem but is actually data starvation.
5. Advance to optimization (weeks 5 to 8) when 30 conversions accumulate. Pause keywords with 50+ clicks and zero conversions. Identify the 20 percent of keywords driving 80 percent of conversions and protect them. Test smart bidding on the highest-volume campaign first.
6. Advance to growth (weeks 9 to 12) when CPA is within 20 percent of target. Increase budget in 10 to 20 percent increments on campaigns below target with impression share under 70 percent. Expand keywords from search term reports. Test new campaign types only when conversion volume supports them.
7. Handle inherited accounts differently. Observe for 1 to 2 weeks before sweeping changes. Document everything (screenshots, exports). Treat historical data from a broken setup as unreliable: it shows what was happening, not what the account can do.
8. Pause the phase clock if tracking breaks. “Fix tracking” becomes the only recommendation. Do not advance phases on incomplete data. Resume only after 48+ consecutive hours of verified accurate tracking.

When to use it

Use this any time you take over a new account, whether brand-new or inherited, and whenever a client expects “results in week one” so you can set expectations honestly. It applies across verticals with calibration: ecommerce moves through phases faster (high data velocity), B2B SaaS slower (extend stabilization to 6 weeks, optimization to 10), local services need call tracking from day one, and high-value verticals require tighter pause thresholds because every click costs $50 to $200. It does not apply when conversion volume is too low for the framework to ever exit stabilization. Below 5 conversions in 4 weeks, escalate explicitly: the standard framework requires data this account is not generating, and pretending otherwise produces false confidence.