Pre-tested copy · send it tonight

Catch Them.

Copy-paste emails you can send your agency that force a real answer on the most common forms of bad PPC work. Each test ships with a decoder for what their typical reply (the direct one, the hedge, and the tell) actually means.

The whole site is built on the same idea: name the tactic, give the question to ask. This page is the operational form of that. Each test maps to one of the red flags. Read the test, copy the email, send it. When the reply lands, look it up against the decoder and you’ll know what you’re working with.

None of this is a gotcha. A good agency answers any of these in two paragraphs without flinching. The test isn’t the question. The test is the response.

Test #2

Broad match consumes most of your budget and there's no negative keyword strategy

Email request, the week before the next monthly meeting. Forces the work to happen before they walk in with a deck.

Subject Search-terms and negatives

Hi [Name], two pulls before next month's deck:

1. The top 50 search terms by spend over the last 90 days, with conversion rate and cost per lead next to each. Pulled from the platform; not aggregated, not editorialised.

2. The negative keyword list as it stands today, with timestamps showing what was added in the last 60 days.

Thanks.

How to read the reply

Direct

Top 50 attached, an audit log of last 60 days' negatives, and a stated broad-match share

They're doing the work. A real answer here usually includes "broad match share moved from X% to Y% over the last quarter" and a list of recent negative-keyword adds (job-intent terms, competitor adjacencies, off-topic clusters). Confirm a written cadence (every two weeks, in writing) and you have a working partnership.
Hedge

“We add negatives when we see issues / monitor at the campaign level”

They aren't actively pruning. "When we see issues" on a smaller account is functionally never. Reply: "When was the last time the negative-keyword list was reviewed end-to-end, and how many entries were added that session?" If the answer is fuzzy, the answer is never.
Tell

“Broad match is the recommended setting / the system is working as designed”

That's a sales line, not a strategy. The system is working as designed, for Google, not for you. If broad match is over 60% of spend with no negative-list cadence, expect to find 30%+ of your spend going to obviously off-topic queries. Audit and a written negatives cadence is the only fix; switching agencies is rarely necessary, but the conversation has to happen.

Read the underlying red flag →

Test #4

Performance Max is running with zero search-term transparency

Email to the account lead, before the next quarterly review.

Subject PMax exclusions and search themes

Hi [Name], before the next review, two things on Performance Max:

1. Brand exclusions: every variant of our company name currently on the brand-exclusion list. Account-level negative keyword list. Category exclusions inside the PMax campaign settings. Screenshots are fine.

2. The most recent search-term insights pull from PMax, with notes on which themes you've excluded since the last pull and which you're flagging.

Thanks.

How to read the reply

Direct

Brand variants listed, negatives shared, search-term insights with last-cycle changes

They've bounded the campaign. PMax is being run as a campaign type, not as a blank cheque. Worth confirming the cadence is monthly going forward and asking for the next review on the calendar.
Hedge

“PMax is fully automated / we monitor at the campaign level”

They haven't done the work. Brand-exclusion lists and account-level negatives have been available to PMax for over a year; "automated" isn't a substitute for setup. Reply with: "Brand-exclusion lists and account-level PMax negatives have been live for a while. Can you confirm what's currently applied?" If the answer is silence or another deflection, escalate.
Tell

“Performance Max is a black box, that data isn’t shared anymore”

Mostly false. Search-term insights and asset-group performance are visible inside the platform. They're either out of date on a year of platform changes, or they don't want you to see what's been bought. Either way, audit time. And on most accounts running PMax this way, 10–30% of "conversions" turn out to be brand cannibalisation already happening through organic.

Read the underlying red flag →

Test #5

The monthly report has no real month-over-month comparison on conversions

Email at any time. Send before the next monthly report goes out, not during the meeting.

Subject Reporting format change

Hi [Name], a small change to the monthly deck format going forward:

Add columns for last month and same-month-last-year on conversions, cost per conversion, conversion rate, and ROAS, per campaign, with percentage deltas. Lead the narrative with the trailing-six-month trend on conversions, calling out any month where the direction is wrong.

Not asking, just letting you know the new format. Will see it in next month's deck.

Thanks.

How to read the reply

Direct

“Done. Trailing-six-month chart added by default. We'll lead with the trend.”

The fact that you had to ask is a small thing. The quick yes is the signal. Most healthy agency relationships will respond exactly this way. It's a 10-minute change to a template. Move on.
Hedge

“We can look into customising the template / our standard framework was built around best-practice metrics”

That's pushback on a 10-minute task. Reply: "It's a column addition; the data is in the platform. What's the timeline to get it into next month's deck?" If they need more than seven days for a column add, the issue isn't the template.
Tell

“Adding too many comparison points may not tell the right story given seasonality”

They're explaining why they don't want to put month-over-month comparisons in writing. There is no other reason to push back on this request. The data tells whatever story it tells; the agency's job is to explain it, not filter it. Audit, and make this format change the test of whether the relationship is salvageable.

Read the underlying red flag →

Test #7

Every flat or declining month is blamed on a "learning period"

Email or Slack to your agency lead, the evening before a check-in. Don't telegraph it on the call.

Subject Quick one before tomorrow

Hi [Name], quick one before our call.

Can you send the date the most recent learning period started, what specific change triggered it, and what the trailing-90-day trend on conversions and CPL looks like, per campaign, not aggregated?

If we're past the documented two-to-three-week window, useful to understand what's actually driving performance now.

Thanks.

How to read the reply

Direct

A specific date, the change that caused it, and a real trend

Working with someone who has the data and knows the account. They won't be defensive about a soft month because they can point to the cause. No further action. This is what the job looks like.
Hedge

“We’re continuing to monitor / patience is critical with Smart Bidding”

Half-true and aimed at ending the conversation. Smart Bidding does need conversion volume, but "patience" without a date or a number is filler. Reply with: "Past two weeks the platform calls it stable. What's the trailing-90 trend? What's the bidding-strategy change history for that period?" If they can't produce either, escalate.
Tell

“The algorithm decides when learning ends / Google handles that automatically”

Wrong. Learning-period status is visible in the campaign UI with a documented exit threshold (typically 50 conversions in 7 days for most strategies). Either they don't know how to find it, or they don't want you to. Either way, this is the moment to book an audit.

Read the underlying red flag →

Test #11

They optimise to the cheapest cost per conversion, ignoring which products actually matter

Email to the agency lead before quarterly planning, board prep, or any conversation about budget allocation.

Subject Conversion values and product-line priority

Hi [Name], one for next planning:

I want to confirm how our product-line priorities are encoded in the campaigns. Specifically:

1. The conversion-action values currently configured in Google Ads: per action, the dollar value next to each, and how each was derived.

2. Whether the bidding strategies are running on Maximise Conversion Value / Target ROAS, or on Maximise Conversions / Target CPA. By campaign.

3. If I told you Product A is twice as strategically valuable as Product B, walk me through what changes at the conversion-value level, the bidding-strategy level, and the budget level.

Thanks.

How to read the reply

Direct

A list of conversion-action values per product, the bidding strategy by campaign, and a model of what doubling A's weight does

They're treating the account as a portfolio. Conversion values reflect deal-size or LTV, the algorithm is bidding on value not just cost, and a strategic shift is something they can model in writing. This is a working partnership. Confirm the values get refreshed at least quarterly with finance.
Hedge

“We optimise to a blended CPL / happy to discuss reallocating”

They don't have values configured at the conversion-action level. Their answer to "what changes if priorities shift" is "we move budget around manually." That can work, but it means the algorithm doesn't see your strategy. It just sees CPL. Ask: "What would it take to get conversion values configured per action and switch the relevant campaigns to value-based bidding?" If the answer is "that's not how we run accounts," that is the answer.
Tell

“Smart Bidding handles prioritisation / the algorithm decides which products perform best”

They're saying the auction outranks your strategy. Either they don't know how value-based bidding works, or they don't want to do the setup work it requires. Your portfolio strategy should not be downstream of an account manager's methodology. Audit, and put numbers on the strategic ranking before the next renewal.

Read the underlying red flag →

Test #13

Google is auto-applying its own recommendations to your account

Email request, any time. The answer is a yes/no plus a screenshot, so there's nowhere to hide on this one. That's what makes it a good test.

Subject Auto-apply settings

Hi [Name], quick one I want on the record:

1. Are any of Google's "auto-apply recommendations" turned on for the account? If so, which categories?

2. A screenshot of the Recommendations > Auto-apply settings page as it stands today.

3. If any are on, who decided to turn them on, and when?

Thanks.

How to read the reply

Direct

“It's all off, here's the screenshot, we apply changes by hand after review”

This is the answer you want. A real shop keeps auto-apply off and treats Google's recommendations as a to-do list to triage, not a set of orders to follow. Ask one more thing to be sure: is the optimization score something they read, or something they try to max out? The first is fine. The second is the problem.
Hedge

“We keep a few low-risk ones on to save time”

Ask which categories, then look them up. "Low-risk" almost always includes things that are not low-risk to you: applying broad match, raising budgets, and "optimize ad rotation," which retires the ad you approved in favour of the one Google prefers. Time the agency saves is not money you save.
Tell

“It keeps our optimization score high / it's Google's recommended best practice”

That's Google's sales pitch repeated back to you. Optimization score measures how closely your account follows Google's suggestions. You can take it to 100% by doing everything Google asks, and that is not the same as doing what's right for your budget. If the score is the goal, the agency has handed its judgement to the company selling you the ads.

Read the underlying red flag →

Sent the email and didn’t like the reply?

Book a free full-service audit this month. I go through the account, listen to your agency call (subject to your recording policies), and deliver a written report on what I find. No charge, no pitch.

Book your free audit