Yes. Before you commit to six or twelve months with an outbound agency, ask for a pilot with a fixed end date, and pay for it. Plan on about 90 days.
A 30-day pilot mostly measures setup: lists, profiles and sending accounts come before a single message leaves, and when new domains and mailboxes have to be bought and warmed, that alone can take a month.
Put three things in writing before it starts: what counts as a qualified meeting, what gets reported every week, and what happens on day 90. And keep the pilot apart from a paid diagnostic, which is a smaller purchase that answers a different question.
Know what a pass proves, too. I wrote it this way in a Journal post on what an outbound program looks like in its opening quarter: "Ninety days is enough to know whether this works for you. It is not enough to know what it will be worth."
Why 30 days only shows you the setup
Lay a pilot over a calendar and the short version falls apart on its own.
For a straightforward company, about 15 days of setup is a realistic range. The list gets built, the sender's profile and the offer get fixed, and the sending accounts get ready.
When the sending side starts from zero, warmup runs alongside the rest of setup. In the post where I walk through domains, mailboxes and warmup, about 15 days from the start to the first live campaign counts as normal, and new mailboxes spend 10 to 14 days of that warming up. Nothing goes out yet, so there is nothing to count.
Ask any agency which of those two your pilot is, and whether its 90 days start at signing or at the first send.
The first month or so of sending mostly produces feedback. You learn which angle earns answers, which industries stay silent, and which job titles reply versus pass you along to somebody else. Everything the campaign does next is built on that.
By month three you have about two months of sending to compare, and the adjustments start to show. More volume goes behind what earned replies, weak sequences get dropped, and the lists get sharper.
So a 30-day pilot ends a week or two after sending starts, or before it starts at all when the mailboxes are new. You would be grading the setup crew.
| Stretch of the pilot | What is going on | What you can fairly judge | What you cannot judge yet |
|---|---|---|---|
| Setup: weeks 1 to 3, up to 5 with new mailboxes | List building, mailbox warmup, profile and offer fixes | The list itself: sample rows and the rules for who gets on it | Replies, meetings |
| First weeks of sending, to about week 8 | First sequences go out and the market reacts | Which segments, angles and titles reply, and how fast replies get handled | Meeting volume at full speed |
| Weeks 9 to 13 | More volume behind what works, weak sequences dropped | Qualified meetings held per month and the direction of the trend | Revenue, unless your sales cycle is shorter than a month |
During setup there is one thing a founder can check without any outbound background: the list being built. Ask for sample rows and the rules for who gets on it. It is the only output of those weeks a non-specialist can fairly judge.
For an outside reference point, an appointment setting buyer's guide from July 2026 puts the usual pilot somewhere between 30 and 90 days, set by how long your deals take and how many people you contact (checked September 2026). That is one vendor's stated position. The calendar above is the case for the long end.
A pilot and a paid diagnostic are different purchases
Part of the confusion around pilots comes from one word covering two things.
A paid diagnostic is a short, one-off piece of custom work before anything is sent. The agency looks at your market, your list, your offer and your profiles, and tells you what it would fix and what it would test first.
A paid pilot is a live campaign. Real people get contacted for weeks, and it is measured on replies, meetings held and what moved forward.
| Paid diagnostic | Paid pilot | |
|---|---|---|
| What you buy | A review and a plan | Live outreach to your market |
| The question it answers | Does this team understand my market and what is broken? | Will my buyers talk to us through this channel? |
| Length | Short and one-off | About 90 days |
| What you hold at the end | A written list of fixes and tests | Meetings held, reply data by segment and message, a trend |
| When it fits | You are unsure the offer or list is ready, or you are choosing between agencies | You have picked an agency and want proof before a long term |
Founders on the selling side who keep getting asked for free audits and sample work get short advice from me: "When someone still wants custom work before signing, charge for a small paid diagnostic."
Flip it when you are the buyer. If an agency is willing to work for weeks without pay to win you, ask why its pipeline leaves room for that. Agencies with proof and demand tend to qualify you back instead of auditioning.
What real campaign numbers look like at 30 and 90 days
The fairest benchmark for a pilot is real campaign numbers over the same kind of window. Here are two, measured differently: a four-month ramp from the start of one engagement, and a single 30-day window from one account.
Both come from client engagements at Smirnov Consulting Group, a Prague-based B2B outbound lead generation agency that runs cold email and LinkedIn campaigns for founder-led B2B companies and books qualified sales calls.
The first is a US advertising agency, with numbers I published in 2023. Before we started, they had three sales calls to show for four months. Then:
| Month | Sales calls booked |
|---|---|
| 1 | 23 |
| 2 | 37 |
| 3 | 55 |
| 4 | 113 |
What moved the curve was ordinary work: more volume behind what had worked in month one, cuts to what had not, better lists and more accounts coming online.
A buyer judging this at day 30 sees 23. A buyer judging it at day 90 sees a line going up and a reason for it. Month four, which a 90-day pilot never reaches, roughly doubled month three. If day 90 says yes, the next contract is paying for that stretch.
The second, published in 2024, covers 30 days on one client's LinkedIn account, backed by email follow-ups. It is the worked example in my post on why a list outweighs the email copy, and here it shows why a 30-day report can mislead even when the numbers look big.
| Stage | Count | Rate |
|---|---|---|
| People contacted | 1,200 | |
| Replies | 289 | about 24% of people contacted |
| Interested in a call | 21 | about 7% of replies |
| Calls booked | 16 | about 1.3% of people contacted |
| Calls held | 12 (the other 4 rescheduled) | 75% of calls booked |
| Contracts signed so far | 3 | 25% of calls held |
289 replies sounds like a campaign on fire. Only 21 of those people wanted a call. A pilot report that leads with replies tells you very little, so ask for the interested, booked and held rows every week.
Then look at "so far": the 3 contracts are the ones signed inside the 30 days. Calls and revenue sit a full sales cycle apart, and outbound does not make that cycle any shorter.
So a first month can produce a number, 23 calls in the ad agency's case, but that was about a fifth of month four, and that is the problem with judging there.
The caveat I attach to fast early results holds for these numbers too: the fastest case you have read, mine included, is the wrong yardstick. Hold your week-four and week-eight numbers against the stages in the timeline table above, and leave somebody's best month out of it.
How to spot a fake pilot
One lead gen agency's guide to structuring a pilot, updated April 2026, estimates that half of the pilots agencies sell are really year-long contracts with a cheaper first quarter, auto-renewal included. That is one agency's estimate, with no data shown for it. The pattern is still easy to test for.
The end date row below is that pattern. The other rows come from the questions and red flags I give founders before they hire a lead gen agency.
| Check | Fair pilot | Warning sign |
|---|---|---|
| End date | It stops on a set date, and carrying on means signing something new | It rolls into a long contract unless you cancel |
| Promises | A plan with the arithmetic behind it | A promised, fixed count of leads or meetings |
| First send | A visible setup stretch and a sample of the list first | Sending starts almost immediately |
| Reporting | Contacted, replied, booked, held and signed, gaps included | Calls booked on their own, or a highlight reel |
| Your side | A plain list of what they need from you: case studies, rebuilt profiles, someone free to take calls | A claim that they handle everything and need nothing from you |
Ask when the first message goes out. An answer of day two is a warning sign. Anyone can send by the end of the week. What counts is what someone sees when they check you out: the profiles, the offer and the website copy that setup is there to fix.
On guarantees, the problem is control. An agency that guarantees a meeting count is promising things it cannot control, such as whether you turn up to the calls, your close rate, and whether your market wants what you sell.
Settle these before the pilot starts
These are questions to settle with any agency, written into the pilot agreement.
- Which segment and which offer, at what volume? One segment, one offer, and a rough monthly number of people contacted. My rule is one offer, one audience, one channel: a message written for everyone reaches nobody in particular.
- What is a qualified meeting? Title, company size, a stated problem. Write it down before the first send, so nobody has to argue about it on day 90.
- What gets reported each week? People contacted, replies, interested, booked, held, qualified. The gaps between those rows belong in the report too.
- What should be true by day 90? Set it against your own sales cycle. If deals take six months to close, a revenue target at day 90 is really a test of your sales cycle.
- What happens if day 90 misses those numbers? Agree now whether you can stop there with no further fee and nothing rolling over.
- How is it priced, and what exactly counts? If any part is paid per lead, get "lead" defined in writing. Per-lead pricing with the word left undefined is on my own red-flag list: you end up buying spreadsheet rows.
- What do you keep if you stop? Lists, sequences, reply history, and who owns the domains and mailboxes the campaign was sent from.
That last question is where a pilot quietly turns into a lock-in. If the lists and sending accounts stay behind, walking away on day 90 costs you the setup weeks all over again.
The agency guide mentioned above builds several of these answers into the pilot terms it recommends (checked September 2026). A buyer can leave with no fee if the agreed criteria are missed, everything the pilot produced is handed over within 14 days of the end, and the fee is fixed or a prorated retainer.
Its reason for avoiding pay-per-lead or pay-per-meeting alone during a pilot is incentive: the agency then has little reason to qualify hard.
If an agency refuses a 90-day pilot and holds out for a six-month minimum, the length is not the whole problem. What an outbound program is worth tends to show around month six. The question is whether you can still stop at day 90, with no further fee, if the numbers agreed for that day miss.
It also helps to agree a midpoint look around day 60. Judge it on the middle row of the timeline table: which segments and angles reply, and how fast replies get handled. A full meeting count comes later.
How any of this is worded in the final agreement is a matter for you, the agency and your lawyer. This is not legal advice.
What a pilot cannot tell you about your side
A pilot tests two companies, not one. If the gap is on your side, the result looks like a bad agency, and the next agency's pilot runs into the same gap.
The engagements that have gone badly for me all went wrong on the client's side. Run these checks on yourself before day one:
- Is the offer already selling? Outbound can scale an offer that already sells. It cannot create demand for one that does not, and a pilot on an unproven offer ends up grading the offer.
- Who is free to take the calls within days? If the first free slot in your calendar is three weeks away, the buyer may have forgotten why they said yes.
- Who answers replies every day? Many replies are neither yes nor no: a question, a pointer to a colleague, a request to come back next quarter. Someone has to own that inbox.
- Will you pause when things get busy? A paused campaign loses its place: half-finished sequences, lists that age, sending accounts that cool down. My position on this one is fixed: "If a company tells me they are likely to pause when things get hectic, I would rather not start."
If any of these is a no, fix it before day one. Fixing it halfway through wastes the weeks already paid for.
Questions founders ask about outbound pilots
How long should an outbound agency pilot run? About 90 days. Setup takes about 15 days for a straightforward company, including when new domains and mailboxes must be bought and warmed alongside it. The first weeks of sending mostly produce feedback, and by month three there are about two months of sending to compare. Judge revenue later if your sales cycle is long.
Is a 30-day pilot enough to judge an outbound agency? Rarely. By day 30, sending has usually only just started, and with new mailboxes it may not have started at all. You can judge the list and how fast replies are handled, but not how many qualified meetings the campaign produces once it is running properly.
What is the difference between a paid pilot and a paid diagnostic? A diagnostic is a short, one-off review of your market, list, offer and profiles, ending in a written plan. A pilot is a live campaign of about 90 days, measured on replies, meetings held and what moved forward.
What should a pilot show you by day 90? Qualified meetings held per month and whether that number is rising, which segments and messages get replies, and what moved to a next step, measured against your own sales cycle. Final revenue usually arrives later.