Playbook 9 of 11

21 Questions That Make Bad Agencies Sweat

Ask these before anyone touches your ad budget. The good ones answer in seconds. The cowboys change the subject. Includes the 5 answers that should end the meeting on the spot.

Chapters
12
Reading time
35 min
Price
Free

What is inside

The first 2 chapters are open. The rest unlock with your work email, on this page.

  1. 1Why the wrong provider looks like the right oneFree
  2. 2The five kill questionsFree
  3. 3Who does the work (questions 1 to 4)With email
  4. 4What counts as a result (questions 5 to 9)With email
  5. 5How it is measured, and how to read their proof (questions 10 to 13)With email
  6. 6What you own when you leave (questions 14 to 17)With email
  7. 7What it really costs (questions 18 to 21)With email
  8. 8The contract terms to insist onWith email
  9. 9The scorecardWith email
  10. 10Do this todayWith email
  11. 11Or have it builtWith email
  12. 12Sources and inspirationWith email

81% of B2B buyers end up dissatisfied with the provider they chose (Forrester, 2024). Of SMBs that quit their marketing agency, 56% leave between month 6 and month 12 (vcita, 2025). That is the exact window a 12-month contract locks.

Every provider sounds the same on the sales call. Big numbers, a slick deck, a founder who "gets it". Then month three arrives and nobody can explain why the calendar is full of people who cannot afford you.

This guide is the opposite of a sales call. It is a buyer-side scorecard for owners doing $30K+ a month who are about to hand someone their ad budget: the questions to ask, the answers that should end the meeting, the contract terms to insist on, how to read a case study, and how to score up to three providers side by side.

One honest note before we start. insaneX sells a done-for-you booking machine. So we wrote a test we have to pass ourselves. Use every question on us. If we fail it, do not hire us.


Why the wrong provider looks like the right one

Smart owners keep hiring bad providers. Not because they are naive. Because the sales call is the one thing every provider has perfected.

The numbers say the problem is common, not personal:

What happenedNumberSource
B2B buyers dissatisfied with the provider they chose81%Forrester, State of Business Buying 2024
SMBs that outsource marketing and then quit or switch agency40%vcita, 2025 SMB Marketing Report, 500 US owners
Share of that churn that lands 6 to 12 months in56%same vcita report
Reason given for leaving: no return on investment60%same vcita report
Top reason clients end agency relationships: unhappy with delivery48%Setup, 6th Marketing Relationship Survey, 400+ respondents
B2B buyers who notice the vendor's website and its sellers say different things69%Gartner, June 2025 release (as reported)

Look at the third row against the contract you were offered. Most of the people who leave do it between month 6 and month 12. A 12-month term is priced for exactly that discovery.

One more honest admission: there is no large, independent study of lead gen agency outcomes. Every "benchmark" in this market comes from someone selling a model, us included. So this guide does not ask you to trust numbers. It teaches you to check them.

The four places every bad provider breaks

Every bad engagement fails in at least one of four places. Each of the 21 questions maps to one.

  1. Definition. What counts as a result, in writing, and who decides when a lead fails it.
  2. Measurement. Whose system holds the number: their dashboard or your CRM.
  3. Ownership. What you keep on the day you leave.
  4. Incentive. What the provider earns more from: your spend, their volume, your term length, or your pipeline.

The payable event ladder

This is the single most useful idea in the guide. Every provider gets judged, and usually paid, on some rung of this ladder. From weakest to strongest:

  1. Impressions, clicks, opens
  2. Form fills ("leads")
  3. Meetings scheduled
  4. Meetings held
  5. Meetings held that pass a written qualification standard
  6. Opportunities created in your CRM
  7. Closed revenue

Then ask one question: "Which rung do you report on, and which rung do you get paid on?"

If either answer sits below rung 4, you have found the gap where your money will go missing. A provider paid on rung 2 is rewarded for loosening the filter. A provider reporting on rung 1 is asking you to applaud the weather.

Rung 7 is a trap too. A provider who guarantees closed revenue does not control your closer, your price or your follow-up. They are either very wrong or planning to blame you in the fine print. The sweet spot is rungs 4 to 6: things the provider genuinely influences and you can verify in your own calendar and CRM.

(The ladder is our restatement of ideas from pay-per-appointment write-ups by DemandZen, Prospeo and LeadHaste. All three are vendors. The ladder is ours.)


The five kill questions

You do not need all 21 to rule out most providers. Five questions end the bad calls early. Ask them in the first 15 minutes. If a provider fails any one, you can be polite, finish the call, and save yourself the next three meetings.

Write down the exact words they use. Words on a sales call become contract words, or they disappear.

Kill question 1: "What is your written definition of a qualified call?"

This is question 5 on the full scorecard and the most important one in this guide.

Good answer: "Here it is. A qualified call is a held call with a decision maker at a company doing $30K+ a month, in one of these industries, with a stated problem we solve and a timeline under 90 days. If one of those is missing, it does not count, and here is what happens when it does not."

Bad answer: "We only send qualified leads. Our targeting is very precise." Or: "We'll define that together during onboarding."

The bad answer means the definition will be written after you pay, by the person who gets paid for volume. "Together during onboarding" sounds collaborative. It means "later, when you have less say".

Your move: ask them to email the definition before you sign. If it does not arrive, that is the answer.

Kill question 2: "Do you bill and report on booked calls or held calls?"

Question 7. This one is pure arithmetic.

Outsourced appointment setting has been modelled at around 42% no-show, against about 8% when the prospect books instantly at the point of the form (Focus Digital, 2026, modelled from published sources, sample not disclosed, so treat it as an illustration).

Worked example. You pay $400 per booked call for 100 booked calls. $40,000.

42% no-show8% no-show
Booked calls100100
Calls that actually happen5892
Real cost per held call$690$435

Same invoice. 34 fewer conversations. The sticker price was never the price.

Good answer: "We report booked, held and qualified separately. No-shows do not count toward anything we are paid on, and here is how we reduce them: instant booking, reminders at set intervals, a confirmation step."

Bad answer: "No-shows are on your side, that's a sales issue." Or a long pause.

Kill question 3: "Whose business owns the ad account and the pixel?"

Question 14. On Meta, an ad account belongs to the business portfolio that created or claimed it, not to whoever logs in. If the account lives in the provider's business, your spend history, your pixel's learning and your audiences live there too. When you leave, they stay.

Good answer: "Yours. We create it in your business portfolio, or you create it and give us partner access. You can remove us any day."

Bad answer: "We run everything through our agency account for efficiency." Or: "We can transfer it if you leave." (Can they? On what terms? In how many days?)

Then check it yourself in 30 seconds (the ownership chapter shows where to click).

Kill question 4: "What is the minimum term, and what are the exit terms?"

Question 19. Remember: 56% of SMB agency churn lands 6 to 12 months in. A 12-month lock-in covers exactly the window where buyers discover the problem.

Good answer: "A first term of around 90 days, with written checkpoints. If we miss the checkpoint numbers, you can leave with 30 days' notice and keep everything."

Bad answer: "Twelve months, but nobody ever wants to leave." Or the reverse extreme: "No contract at all, cancel anytime," with no plan for setup or testing.

Month-to-month is not automatically safe. No minimum often means no investment in setup or testing. You want a short first term with exits written in.

Kill question 5: "How do you make more money: when I spend more, when you book more, when I stay longer, or when my pipeline grows?"

Question 21. There is always an answer. The honest providers say it out loud.

If they earn more when...They are quietly rewarded for...
You spend more (percentage of spend)Raising your budget, not your results
They book more (pay per lead or per booked call)Loosening the filter
You stay longer (long lock-in)Surviving month 6, not winning it
Your pipeline grows (paid on held, qualified calls or pipeline in your CRM)The thing you actually want

Percentage of spend is the quiet one. A 15% fee and a $3,000 flat fee cost the same at $20,000 a month of ad spend. At $60,000 a month, the percentage model costs $9,000 a month for roughly the same work. Nothing wrong with paying more for more work. Just know which one you are paying for.

Even big advertisers struggle here. The ANA found 41% of advertisers used performance incentives in agency pay in 2022, down from 61% in 2013, and the most common trigger was a subjective review. If big-brand procurement teams cannot pin incentives to hard outcomes, a handshake on a sales call will not.

The kill rule

Any bad answer to one of these five ends the evaluation, no matter how good the rest sounds. A charming founder with great case studies who owns your ad account and won't define a qualified call is still a bad hire. Charm is not a contract term.

Those five alone will clear most of your shortlist. The rest of the guide covers all 21 questions, 10 contract clauses, two reference call scripts and the scorecard.

Rather have it built and running?

We'll Build You A B2B Booking Machine That Lands 5-7 Qualified Sales Calls On Your Calendar Every Single Day.

Ads, funnel, CRM, and email/SMS follow-up, built and live in 14 days.

For B2B companies doing $30K+ a month.

Book a strategy call