Buy a Paid Pilot Before You Buy a Twelve Month Retainer
The contract in front of you commits you for twelve months and lets the agency leave on thirty days notice. A twelve month retainer asks you to pay for trust you do not have yet.
The standard offer, and who carries the risk in it
A long committed term with a short notice window is not symmetric: you are committed for the whole ramp, the agency only for its notice period. It bites hardest early, because the first two months are setup: an audit, access collection, tone and brand work, a calendar, and modest published output. Real work, not a stall, and the stretch with the least evidence of results.
Agencies want the year for staffing, not as a trick: they assign named people against committed revenue. A wrong fit shows in month two, you pay through month three, and the setup work is hard to use without them.
What a paid pilot is, and what it is not
A paid pilot is a fixed term of two to three months at or near the real monthly rate, with a named deliverable, full access from day one, and a written decision on an agreed date.
It is not a discounted trial: paying less invites junior staffing, so you test a team you will never be assigned. It is not a spec pitch or free audit, which are unpaid samples. Nor an open ended monthly deal, which forces no decision.
The good version ends in a strategy document, a live calendar, published work, and a decision meeting. The bad version is cheap, vague help with no artefact, no access, no date.
The five terms the pilot has to state in writing
- Term and price. The exact months, the real monthly fee, and a start date tied to the day access is granted, not the day you sign. A slow approval should not eat a tenth of the pilot.
- Named deliverable. One or two artefacts that survive whatever the results: a strategy document, a calendar for a stated period, and output as a number of pieces, not “regular content”. A scope question, so reading the scope of work applies.
- Access granted. The accounts, ad accounts, analytics, brand assets, and named approvers, on day one. A pilot you starve fails for your reasons: what you owe in the first thirty days.
- Continuation criteria. The written list of what would make you continue, agreed before work starts. Criteria assembled in month three are a justification, not a test.
- Exit and ownership. Who holds the accounts, who owns drafts and finished assets, what transfers if you stop. Covered in what to check in an agency contract and who owns your social accounts. Commercial terms, not settled law, so have them reviewed.
Defining success when the results are not in yet
Concede the objection, which is usually right: social outcomes do not resolve in eight weeks, since audiences compound and paid campaigns need data before performance settles. So split the evidence. Performance evidence is what the work achieved, which a pilot cannot settle honestly. Process evidence is whether this team can be relied on to produce work at all, which a pilot can.
Process criteria are observable. Did they ship the agreed volume on the agreed dates. Did drafts need edits to substance, meaning the business was not understood, or only to detail. Did they ask for real context, your margins, your sales cycle, the products you want to move. Did reporting explain causes or list numbers. Did the pitch’s senior person appear in the work or vanish after week two.
Turnaround, consistency and draft quality are fair to judge this early. Revenue attribution and campaign cost per acquisition are too noisy. A fair failure is competent work you still decline.
Asking for it without sounding like a tyre kicker
Wanting to spend less is what makes a buyer look uncommitted, so remove that signal. Full rate, approved budget, a decision date, criteria offered up front. A template to adapt, not a real exchange:
“We want to work with you and we will not sign twelve months first. We can sign three months today at your standard rate, scoped to the strategy document, calendar and volume you proposed. Budget is approved. You will have our criteria before you start, we decide in week ten, and we sign the annual term at the rate we agree now.”
Offer something back, since they carry the staffing risk: written intent to sign the annual term on stated criteria, the retainer rate fixed now, or a non refundable fee.
A refusal is information. Ask which part is the problem: staffing economics is answerable with a longer pilot or a later start, refusing to be measured is not. Be willing to walk, and browse vetted agencies or list your agency free.
The decision at the end of the pilot
Book the decision meeting before the pilot starts, two or three weeks before the end so an exit is orderly. Decide in the final week and the default becomes extension.
Score against the written criteria and across the whole term, since the closing weeks are an agency’s strongest. Three outcomes are legitimate: sign the longer term at the rate already agreed, extend once for a stated reason with a new date, or stop. An unlimited extension is a fourth outcome dressed as the second.
If you sign, the criteria become the reporting cadence, so the pilot leaves a way to manage the relationship. Term length is separate from budget: see the four budget bands.
FAQ
How long should a paid pilot be?
Two to three months. It needs a full onboarding plus a stretch of steady output. One month is mostly a test of onboarding speed, since setup consumes it.
Should I ask for a discount on a pilot?
No. Agencies staff against what an engagement earns, so a discounted pilot gets junior staffing and you evaluate a team you will never be assigned. Pay the rate and negotiate the term.
Is a pilot fair to the agency?
Yes, when the fee is the standard one, the access and approved spend arrive as promised, and the end date and the test are written down before anyone starts. It is unfair when access sits behind an internal approver, no decision date exists, or the bar moves once the work is in.
