I added questions to my Meta lead form and volume dropped. Did quality actually improve?

Friction filters for effort, not for money

Every extra field removes people. The question is who it removes. A field asking whether the person owns or rents removes renters, and for most home services renters cannot buy. That is a real screen. A field asking for a preferred appointment window removes people who are on a phone in a parking lot and do not want to think about it, which includes plenty of buyers.

So the useful mental model is not more friction equals better leads. It is: does this question correlate with the outcome I want? If you cannot state why a question predicts booking, it is probably just costing you volume.

Measure at the outcome, not at the form

Cost per lead will always look worse after adding friction, and cost per lead is the wrong scoreboard here. The number that decides it is cost per booked job: spend divided by jobs actually booked from that source in the same period.

That requires the lead to be followed all the way into the operational system, which is why this test is impossible to run honestly without revenue-level reporting. If your reporting stops at the ad platform, adding form questions is a decision you will make on vibes.

Run it as a real comparison, not a switch

• Run both forms simultaneously. Sequential tests confound the change with weather, season and competitor spend, all of which move local service demand.
• Give it enough booked jobs, not enough leads. The metric you are judging is downstream and rarer, so the test needs more time than a lead-count test would.
• Hold creative and audience constant. Changing the form and the hook at once produces a result you cannot attribute to either.

The version most operators land on

A common resting point is a short form with one hard qualifier and no soft ones. Ownership status, property type, or a yes/no on whether the problem is urgent. Everything else that a dispatcher needs gets collected on the follow-up call, because a human on the phone can gather more in thirty seconds than a form can in five fields.

That shifts the burden from the form to the call, which is fine if the call handling is strong. If it is not, adding form questions is treating a symptom. Call analysis will tell you which problem you actually have, and the marketing side can then be tuned against the real constraint.