The 90-day framework with success criteria, exit terms, and the differences between a real pilot and a contract dressed up as one.
By the Launch Leads team · 6 min read · Updated April 2026
Structure a real 90-day lead gen pilot with four parts that the contract makes legally enforceable: a 90-day window (long enough to read signal, short enough that the agency can’t run out the clock), a defined ICP segment (one segment, not your full ICP), written success criteria with specific numbers (volume target, quality bar, sourced pipeline target), and clean exit terms (no auto-renewal, pro-rated refund, CSV data handover within 14 days). Agencies who push back on any of the four are telling you the pilot is theater.
The agency told you they “of course” do pilots — and now you’re trying to figure out what a real pilot actually looks like.
Half of the “pilots” lead gen agencies run aren’t pilots. They’re a discount on the first three months of a 12-month contract, with the same auto-renewal and the same lock-in dressed up as a trial. The pilot pricing makes you feel like you have an out. The pilot structure ensures you don’t.
A real pilot is short, narrow, defined, and exitable. The structure is the whole point.
The thesis: a real pilot has four parts that the contract makes legally enforceable: a 90-day window, a defined ICP segment, written success criteria with specific numbers, and a clean exit if the criteria aren’t met. Agencies who push back on any of the four are telling you the pilot is theater.
What we’ve learned across 1,000+ B2B engagements
The answer: A real pilot is real if it has all four parts: a 90-day window (60–90 days is enough for any modern B2B lead gen playbook to show signal — shorter doesn’t read, longer lets the agency run out the clock), a defined ICP segment (one narrow segment that’s testable), written success criteria (specific numbers in the SOW, not adjectives), and clean exit terms (no auto-renewal, pro-rated refund, data handover). Missing any of the four turns the engagement into a discounted contract.
A pilot is real if and only if it has all four:
If any of these four are missing, the engagement is a discounted contract pretending to be a pilot.
The answer: Three numbers, agreed in writing before kickoff: volume (“X qualified meetings per month, where qualified means [specific definition]”), quality bar (“meeting acceptance rate of at least Y%, measured weekly, with a 5-business-day rejection window”), pipeline outcome (“$Z in sourced opportunities advanced to [specific stage] by day 90”). Each criterion has a number, a definition, and a measurement window. No adjectives. No “high quality.” No “strong impact.”
Three numbers, agreed in writing, before kickoff:
Numbers vary by category, motion, and ICP, so don’t anchor on the example. The structure matters: each criterion has a number, a definition, and a measurement window. No adjectives. No “high-quality.” No “strong pipeline impact.”
If the agency proposes criteria that contain only volume — “we’ll deliver 30 meetings in 90 days” — that’s a volume guarantee, not success criteria. Volume without quality is the original red flag this content series exists to call out.
“We were incredibly impressed with how fast it was from the time that we originally contracted Launch to the time that they were up and running making calls — literally in a matter of days.”
— Shauna Dickerson, Director of Marketing, Corda
The answer: Six lines of exit language must be in the SOW: pilot ends at day 90 by default with no auto-renewal, continuation requires explicit re-signature (not silence), buyer can exit with no fee if success criteria aren’t met, unmet portion of the qualified-meeting target is pro-rated and refunded, all data and recordings are buyer property delivered in CSV within 14 days, and post-pilot contract terms are negotiated separately (not auto-attached to the pilot).
Six lines in the SOW. All six need to be there:
The most common version that fails this test: a “pilot” that has a discount on months 1–3 but auto-rolls into a 12-month contract at standard pricing. Two of the six lines are missing, and the engagement becomes hard to leave the moment day 90 passes.
The answer: Expect pushback on three of the six exit lines. “We need a 6-month minimum to see results” means the playbook isn’t strong enough to show signal in 90 days. “We can’t pro-rate the meeting target” means they’re not confident they’ll hit it (pro-rating costs the agency nothing if they hit, only money if they miss). “Data handover is standard at end of full engagement, not pilot” means they want you to feel switching costs at day 90 even if you exit.
Expect pushback on three of the six lines. The pushback is informative.
“We need a 6-month minimum to see results.” Translation: the playbook isn’t strong enough to show signal in 90 days. Real agencies have tightened their playbooks to where 60–90 days is enough. A 6-month minimum is the agency protecting against weak early performance.
“We can’t pro-rate the meeting target.” Translation: they’re not confident they’ll hit it. Pro-rating is asymmetric — it costs the agency nothing if they hit, costs them money only if they miss. Refusing it tells you they expect to miss.
“Data hand-over is standard at end of full engagement, not pilot.” Translation: they want you to feel switching costs at day 90 even if you exit. Push back. The data is yours; the format is CSV; the timing is 14 days; this is non-negotiable.
Agencies who agree to all six lines without redlines are agencies who have run real pilots before. Agencies who fight all six are showing you what the engagement will feel like.
“The Launch teams, I think we maybe spent 2 hours, maybe 3 hours tops over two days, and they were off and running. We were getting appointment notifications within an hour after our first training. One deal from that first show more than paid for our investment in Launch.”
— Eric Flynn, CEO, Treehouse Interactive
The answer: Three pilot checkpoints: day 30 — ramp signal (first meetings, first acceptance/rejection data, first sequence iterations — grade on whether the agency is iterating, not on outcomes yet), day 60 — trajectory check (acceptance rates rising, sourced-opportunity volume building — flat trajectory means call it now), day 90 — the decision (success criteria hit means negotiate the post-pilot contract; missed means exit cleanly). Don’t extend “to give it more time” — 30 more days won’t fix what 90 didn’t.
The 90-day pilot has three checkpoints:
Don’t extend the pilot to “give it more time.” If 90 days wasn’t enough, 30 more won’t fix it. The pilot exists exactly so you don’t keep paying past the point where the relationship has shown you what it is.
For most B2B lead gen, yes. By day 60 you should see early acceptance signal; by day 90 you should see sourced pipeline. If you don’t, more time usually doesn’t fix it — the playbook isn’t right for your motion.
Two common shapes: a fixed pilot fee (paid upfront or in two installments) or a pro-rated retainer with a credit-back clause. Avoid PPL/PPM-only pricing on a pilot — the agency loses incentive to qualify hard, and the ICP-narrowing protection breaks down.
Generally no. Splits the buyer-side attention, splits the agency-side data, dilutes the comparison. Better to run a structured evaluation, pick one finalist, run a 90-day pilot, and have the runner-up on standby if the pilot fails.
Yes — if the contract terms (length, pricing, notice windows) were negotiated separately during the pilot. Don’t auto-attach pilot terms to a 12-month contract; that’s the trap. Renegotiate the full contract before day 91.
That’s not a pilot. Either rewrite the SOW to include the six exit lines above, or treat the engagement as a normal contract — and expect to be locked in at day 91 regardless of what the agency calls it.
Get on a call with our team. We’ll walk through the SOW the agency is proposing, mark what’s missing, and tell you what to negotiate before kickoff.
WHAT YOU GET
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.