You're looking at the prep tool
I'd run my book with.

The Room

A VP of Performance who owns blended CAC, and a Head of Measurement with the veto. Interested, unconvinced, better at math than you.

Objection One

They say

"Attribution is unproven and the reporting is thinner than Google's."

What's true

It is, today.

The counter

Design the test on their incrementality framework, not our reporting.

Next step

Their measurement lead in meeting one, not meeting three.

Objection Two

They say

"By the time the ad shows up, the model already told the traveler what to book."

What's true

Sponsored slots render after the answer. Published design.

The counter

In travel the answer is a shortlist, and a shortlist is where a contender wins.

Next step

Build around comparison queries, not single-answer ones.

Objection Three

They say

"Our budgets are committed to Google and Meta."

What's true

Nobody unwinds core spend for an early platform.

The counter

The ask is experimental budget with a proof bar, not a reallocation.

Next step

Size it so finance can approve without touching the annual plan.

Test Structure

Geo-holdout or cohort test. Incremental bookings and cohort CPA vs blended CAC. Scale gate agreed in writing.

The Close

A 90-day cohort on their framework, exec review at day 45. Small enough for a yes, structured enough to scale.

The Room

A VP of Media and an eCommerce lead, with brand and legal one door down. Curious, brand-protective.

Objection One

They say

"An AI conversation is a brand-safety risk we can't control."

What's true

The bar should be high.

The counter

Ads are labeled and separate from answers; sensitive-conversation limits are published policy.

Next step

Do the policy walkthrough before they ask.

Objection Two

They say

"What if the model recommends against us and our ad sits under it?"

What's true

The answer is earned, not bought. As it should be.

The counter

Show up in comparison moments in categories you already win on merit.

Next step

The ad amplifies. It never argues.

Objection Three

They say

"Early click-through rates look weak next to search."

What's true

Early CTRs trail Google's. On record.

The counter

Measure assisted basket size and new-to-brand rate, not last-click.

Next step

Agree the metric set before launch, or the retro writes itself.

Test Structure

Assisted-journey cohort vs a paid-social lookalike: new-customer CPA, basket size, loyalty signups.

The Close

One category, 60 days, brand team in the room.

The Room

A VP of Growth plus finance, at a company that runs its own ads platform. They'll out-math a weak pitch in ten minutes.

Objection One

They say

"Your CTR is lower than what we get on Google."

What's true

On early benchmarks, yes.

The counter

Judge the surface on CPA and incrementality, not CTR.

Next step

Offer the geo-holdout on day one.

Objection Two

They say

"Branded search already captures this demand."

What's true

Some of it, at a rising price.

The counter

A new surface is the cleanest read on what's truly incremental.

Next step

Frame it as the cannibalization audit finance already wants.

Objection Three

They say

"The tooling is rudimentary. Very few levers."

What's true

Fewer levers than Google Ads. No pretending.

The counter

The window is before the tooling matures and the auction fills.

Next step

Their product feedback becomes part of the deal.

Test Structure

Geo-holdout, two matched metros: new-customer CPA, 90-day frequency, measured cannibalization.

The Close

Six weeks, two metros, read together weekly. If the holdout works, scaling is a spreadsheet conversation.

The Room

Compliance first, CMO second, growth third. That ordering is the point.

Objection One

They say

"We can't risk a regulatory headline for an experimental channel."

What's true

Completely. No test is worth an enforcement letter.

The counter

Policy before commerce: cleared products, approved claims, then spend.

Next step

First meeting is compliance-to-policy, not sales-to-marketing.

Objection Two

They say

"AI answers about money are a minefield."

What's true

Money deserves the highest bar on any surface.

The counter

Answer-ad separation beats any keyword blocklist they run today.

Next step

Policy docs in their compliance file before the pitch deck.

Objection Three

They say

"The measurement is thin, and our audit trail needs more."

What's true

Reporting is early. Regulated advertisers need more.

The counter

Define the audit trail together; their requirements become the template.

Next step

Route them to the ads team as product feedback.

Test Structure

Cleared-scope pilot, one product line. Metrics set after scope, not before.

The Close

Two quarters: policy now, pilot next. How I built Google's regulated book from zero.