Skip to main content
Buyer’s Guide · Evaluation

How to Do Reference Checks on a Lead Gen Agency

7 questions that sort finalists — plus how to find non-curated references the agency didn’t hand-pick.

By the Launch Leads team · 6 min read · Updated April 2026

Run reference checks on a lead gen agency by finding two non-curated references yourself (via LinkedIn or the agency’s published case studies), adding the agency’s curated reference for three calls per finalist, running each call in 25 minutes, and asking seven questions in priority order: what was the hardest part of the engagement, how did the agency handle it, what did they do better than you expected, what worse, was the team you started with the team that delivered six months in, what did the dashboard hide, and would you sign with them again knowing what you know now.

You’ve got two finalists and a “reference list” from each. Six logos per agency. Names and emails. You’re trying to figure out how to actually use them.

Most reference calls don’t sort finalists because most reference calls are run wrong. You take the list the agency provides, you book a 30-minute call, you ask “how was working with them?” and the reference says “great.” You hang up and you’re no closer to a decision than you were before.

The reference checks that actually change finalists are not the ones the agency hands you. They’re the ones you find yourself.

The thesis: the highest-signal references are the ones the agency didn’t curate. Find two via LinkedIn or the agency’s published case studies, take the agency’s curated list as a third, ask the seven questions below, and let one finalist drop out on the merits.

What we’ve learned across 1,000+ B2B engagements

76,000+
Appointments set
26,000+
Sales closed
$3B+
Revenue sourced

The 7-question reference-check script — 25 minutes per call, three references per finalist (one curated and two found yourself), with the question, what each tests for, and what to listen for in red

1. Why curated reference lists are mostly noise

The answer: Curated reference lists are noise. The agency’s reference list is six best clients, hand-picked, primed to say nice things — three are friends of the founder, one is a current customer who’ll lose service if they badmouth the agency. They tell you the agency has at least six happy customers (useful baseline), but they don’t sort finalists. The references that change finalists are the ones you find yourself.

The agency’s reference list is the agency’s six best clients, hand-picked, primed to say nice things. Three of them are personal friends of the founder. One was a five-figure pilot that ended six months ago and they don’t quite remember the details. One is a current customer who’ll lose service if they badmouth the agency.

This is not a trick — it’s just what reference lists are. They exist for the same reason every company has a careers page that says “we love our employees.” But they don’t sort finalists.

A reference call with even one curated reference is fine — the agency should be able to point you at one happy customer. A reference call with all curated references tells you only that the agency has at least six happy customers. Useful baseline. Not a tiebreaker.

2. How to find non-curated references

The answer: Find non-curated references in three places: LinkedIn (search the agency’s name, filter by People, contact employees of brand-name clients who’ve publicly engaged with the agency’s posts), published case studies (every named client on the agency’s site has a contact reachable through LinkedIn), and industry communities (Slack groups, Reddit, trade-press articles). For each agency, find two non-curated plus the agency’s curated reference for three calls per finalist.

Three places to look:

  • LinkedIn. Search the agency’s name in the search bar, filter by “People.” You’ll see employees, but also the brand-name clients who’ve publicly engaged with the agency’s posts. Their employees who’d recognize the agency are reachable.
  • Published case studies. Every case study on the agency’s site has a named client. The case study has the names; LinkedIn has the contact info. Two minutes of searching gets you the right person.
  • Industry communities. Slack groups, Reddit threads, and trade-press articles where buyers compare lead gen agencies. The buyers who chose this agency are usually willing to talk; the buyers who chose a competitor are even more willing.

For each agency, find two non-curated references. Add the agency’s curated reference to make three. Three calls per finalist, three finalists, that’s nine calls — call it five hours of work for what might be a $200K decision. The math is on your side.

“They’ll come through with what they say they can deliver, and they know their stuff. They’re able to get a foot in the door and get with the decision-makers at the companies you’re trying to get in with.”

— Dave Bascom, CEO, SEO.com

3. Running the call — agenda and timing

The answer: Run the call in 25 minutes, not 30. Minute 1: thank them and frame the call. Minutes 2–15: the seven questions. Minutes 15–22: open-ended (“Anything I should have asked?”). Minutes 22–25: close with “may I email you one follow-up question if it comes up?” — references who wouldn’t take a follow-up don’t really endorse the agency. Don’t share your specific situation in detail; ask how the agency handled their problem.

Twenty-five minutes, not thirty. Buyers are doing you a favor; respect their time.

The structure:

  • Minute 1: thank them, frame the call. “We’re evaluating [Agency] for [type of engagement] and would value 25 minutes of your perspective. There are no wrong answers — we’re trying to make a real decision.”
  • Minutes 2 to 15: the seven questions in the next section.
  • Minutes 15 to 22: open-ended. “Anything I should have asked?”
  • Minutes 22 to 25: close. “If you don’t mind, may I email you one follow-up question if it comes up?” (This makes them honest — if they wouldn’t take a follow-up, they don’t really endorse the agency.)

Don’t share your specific situation in detail. Asking the reference how the agency handled their problem is more useful than asking them to predict how the agency will handle yours.

4. The seven questions that sort finalists

The answer: Seven questions in priority order: (1) what was the hardest part of the engagement, (2) how did the agency handle that hard part, (3) what did they do better than expected, (4) what did they do worse than expected, (5) was the team you started with the team that delivered six months in, (6) how was the dashboard — what did it tell you, what did it hide, (7) would you sign with them again knowing what you know now. The decisive question is the seventh; “yes, but…” is the strongest signal.

In order of priority:

  1. “What was the hardest part of the engagement?” Tests whether the relationship was real. Every real engagement has a hard part. References who say “nothing was hard” either weren’t paying attention or are protecting the agency.
  2. “How did the agency handle that hard part?” Tells you the agency’s failure mode. Did they iterate? Did they go quiet? Did the senior people show up or did the junior people stop responding?
  3. “What’s something the agency did better than you expected?” Tests for genuine enthusiasm. References who can name something specific are different from references who say “they did everything well.”
  4. “What’s something they did worse than you expected?” The mirror image. Real references can name one thing. Curated references can’t.
  5. “Was the team you started with the team that delivered six months in?” Tests for bait-and-switch. Pitch teams sell, delivery teams execute, the gap is often the problem.
  6. “How was the dashboard / reporting? What did it tell you, what did it hide?” Real references will name something the dashboard hid. Curated references will say “great reporting.”
  7. “Would you sign with them again, knowing what you know now?” The decisive question. A confident yes is a strong signal. A “yes, but…” is a stronger signal — that’s where the real information lives. A “no” is rarer and you should ask why before you weight it.

5. What to listen for besides the words

The answer: The words are half the signal. The other half: pause length before the answer (confident references answer within a beat; editing references hesitate), specificity vs. generality (names, numbers, dates beat “they were great”), voluntary detail (references who keep going past the question are enthusiastic), and tone shift on hard questions (energy drops on questions 4 and 6 mean the engagement had texture; energy stays the same means autopilot).

The words are half the signal. The other half:

  • Pause length before the answer. A confident reference answers within a beat. A reference who has to think before answering is editing.
  • Specificity vs. generality. Names, numbers, dates beat “they were great” every time. Generality is the language of references who don’t quite remember.
  • Voluntary detail. A reference who keeps going past the question is enthusiastic. A reference who answers the question and stops is being polite.
  • Tone shift on hard questions. If the energy drops on questions 4 and 6, the engagement had texture. If the energy stays the same, the reference is on autopilot.

Take notes. Compare across the three calls per finalist. The patterns emerge in the comparison, not in any single call.

6. What to do with what you hear

The answer: After all nine calls, write one paragraph per finalist answering: what’s the consistent theme, what’s the consistent gap, which finalist’s hard-part story matches your situation most closely, where’s the bait-and-switch risk. Most decisions sort themselves at this stage. The finalist whose references can’t quite name what was hard usually drops; the finalist whose references mention the same delivery person by name twice usually advances.

After all nine calls, write one paragraph per finalist:

  • What’s the consistent theme across their references?
  • What’s the consistent gap?
  • Which finalist’s hard-part story matches your situation most closely?
  • Where’s the bait-and-switch risk?

Most decisions sort themselves at this stage. The finalist whose references can’t quite name what was hard is usually the one to drop. The finalist whose references mention the same delivery person by name twice is usually the one to advance.

“They get up and running fast, the people they have are talented, they’re experienced. They’re people who understand how to sell to the types of customers that we deal with.”

— Eric Flynn, CEO, Treehouse Interactive

For the broader evaluation framework, see How to Choose a Lead Generation Company. For the questions to ask the agency directly (different list, different purpose), see 12 questions every lead gen vendor hopes you don’t ask.

Frequently asked questions

How many references should I call per finalist?

Three. One curated (from the agency), two non-curated (LinkedIn or case studies). Fewer than three and you don’t see the pattern; more than three is diminishing returns.

Should I email the reference questions in advance?

No. Spontaneous answers are more useful than rehearsed ones. The exception: if you’re asking for very specific data (like “what was your meeting acceptance rate?”), give them 24 hours to look it up.

What if the agency only provides one reference?

That’s an answer. Real agencies have at least three clients willing to take a 25-minute call. If they can only produce one, the issue is upstream — find your own non-curated references and weight them more heavily.

Can I record the calls?

Ask. Most references will say yes. Record what they say verbatim — your memory of the call will compress the most useful nuances within 48 hours.

What if the references contradict each other?

That’s the most useful outcome. Three references saying the same thing tells you the surface story. Three references saying different things tells you what the engagement actually feels like. Compare the two non-curated references to the curated one — usually the daylight is in that comparison.

Free needs assessment · No commitment

Want help running reference calls on the agencies you’re evaluating?

Get on a call with our team. We’ll talk through your finalists, help you find non-curated references, and walk through the seven questions for your specific situation.

WHAT YOU GET

  • → 30-minute call with a Launch Leads pipeline strategist
  • → Talk through your finalists and how to find non-curated references
  • → A look at how Launch Leads would answer the same seven questions

Get a free needs assessment →

If we’re not the right fit for what you need, we’ll say so on the call.

Launch Leads is a B2B lead generation company that has set 76,000+ appointments and sourced over $3B in client revenue across 1,000+ engagements. We focus on multi-channel outbound, real-person outreach, and pipeline outcomes — not activity metrics.

Schedule Discovery Call