Start with the question the pilot must answer
Before you talk volume, write down the one decision the pilot is meant to support. For most buyers it is some version of: can this source produce calls my agents convert at a cost per acquisition I can live with? Everything else, including states, caps and hours, should be set so that question can be answered cleanly.
If you try to test five verticals, twenty states and three pricing models at once, you will end the pilot with a spreadsheet full of small numbers and no decision. Narrow first, then widen.
Pick states you can actually cover
Match the pilot footprint to the states where you have licensed agents on the floor during the pilot hours, not every state you are appointed in. A call routed to a state where only one agent is licensed, and that agent is on another call, is a call you paid for and could not work.
- List the states where you have at least two licensed agents available during pilot hours
- Exclude states where carrier appetite or underwriting makes closing unlikely
- Keep the list stable for the whole pilot so results stay comparable
Set caps from agent capacity, not ambition
Daily and hourly caps protect your answer rate. Work backwards from how many agents will be ready to take calls, how long an average call runs in your vertical, and how much buffer you want for wrap-up. A cap that keeps agents busy but not overwhelmed will tell you more than a cap that floods the queue.
Agree the billable-call definition in writing
Pay per call programs live or die on the billable-call definition. Before launch, agree exactly what makes a call billable: the qualification criteria confirmed on the call, the minimum duration, and what happens with duplicates, wrong states and out-of-hours calls. Put it in the program agreement so both sides read the same numbers.
- Qualification criteria, stated plainly (for example age range, state, coverage intent)
- Minimum call duration that counts as billable
- How duplicate callers inside a set window are treated
- The dispute window and the evidence used, such as call recordings where agreed
Choose dayparts that match your floor
Route calls only inside hours when your team is fully staffed, in the consumer’s time zone. Evening and weekend demand can be strong in some verticals, but only test it once your core hours are working. Changing hours mid-pilot muddies the results.
Decide your success metrics before the first call
Agree the handful of numbers you will judge the pilot on and how you will collect them. Dispositions coming back from your team are what let the source improve qualification during the pilot, so set up the feedback loop on day one.
- Answer rate: calls answered by a licensed agent ÷ calls delivered
- Billable rate: billable calls ÷ calls delivered
- Close rate on billable calls, tracked from your CRM
- Cost per acquisition: total spend ÷ policies written
- Top disposition reasons for calls that did not close
Give it enough volume and enough time
A handful of calls will not tell you much. Plan the pilot so you reach a meaningful number of billable calls across a few weeks, including at least two full weekly cycles, so a slow Monday or a strong Friday does not decide the outcome for you.
Plan the scale-up before you need it
If the pilot works, the next question is how fast volume can grow without breaking quality. Agree up front how caps will step up, which states come next, and what would make either side pause. That way a good result turns into a program instead of a renegotiation.