How to Choose a Lead Generation Agency

What you are actually buying

Lead generation agencies sell three distinct things and frequently blur them in the same proposal: demand creation (making a market aware of a problem), demand capture (converting existing intent), and outbound execution (contacting a list you agree on).

Most agencies calling themselves lead generation are doing the third. That is a legitimate and useful service, and it will not work if the underlying issue is that nobody in your market knows the problem exists — see demand generation vs lead generation. Establishing which you are buying prevents most disappointment.

Pricing models and what each incentivises

Pay-per-meeting deserves particular scrutiny because it sounds well aligned and is not. A meeting is booked when someone accepts a calendar invite, which is a much lower bar than being a genuine prospect — and the agency's margin improves with every marginal acceptance.

The clause that matters most

An uncapped rejection right with mandatory reasons. Every lead or meeting you reject comes off the invoice, and every rejection carries a reason from a fixed list.

Two things follow. The agency's incentive realigns immediately, because loosening qualification now costs them. And within a month you have a dataset showing whether the problem is targeting, timing, or your own follow-up — which is useful regardless of how the relationship ends.

• No rejection cap. Agencies routinely propose 10–15%. If their qualification is good, the cap costs them nothing, so the resistance is informative.
• Reasons from a fixed list — wrong size, wrong industry, wrong role, no budget, no timeline, already a customer, competitor.
• A defined review window — five business days, so rejections are timely rather than retrospective.
• Monthly reason review together. The patterns should change what they do, and if they do not, that is the signal to leave.

The domain question

Ask which domain outreach is sent from and who owns the sending infrastructure. This is rarely in a proposal and it matters enormously.

An agency sending from your primary domain is spending an asset that takes months to rebuild if they burn it. An agency sending from lookalike domains they own is safer for you, and you should still ask about their warming practice, volume per mailbox, and list hygiene — because a shared reputation across their client base can affect your deliverability too.

Structuring a pilot

1. 01
   One segment, three months

   Long enough to see a full sales cycle. Anything shorter measures message quality rather than whether the programme works, and both parties will draw the wrong conclusion.

2. 02
   You provide the ICP and the definition

   Their targeting against your criteria. If they cannot work to your definition, that is the finding, and it is cheaper to learn in a pilot.

3. 03
   Agree the success measure before starting

   Qualified opportunities created, not meetings booked. Write the number and the date into the pilot agreement.

4. 04
   Instrument it separately

   Their leads tagged distinctly in your CRM so you can compare conversion against your other sources honestly.

5. 05
   Review rejection reasons at week four

   Early enough to correct targeting, late enough to have data. This session tells you more about the agency than the pilot's final numbers.

Red flags

When an agency is the wrong answer

Three cases. When your ICP is not yet defined, an agency will define it for you by default, and their definition will be broader than yours would have been. When your follow-up is slow — an agency delivering leads into a system with a 30-hour median response time is money spent to feed a leak, which is a routing problem you should fix first.

And when the real constraint is that nobody has heard of you. Outbound into a market with no awareness converts poorly regardless of execution quality, and the honest answer is that the budget belongs in demand creation for two quarters first.