Dayparting with Smart Bidding: still worth setting ad schedules?

Cutting ads overnight and pushing budget into business hours is old advice, and with Target CPA or Maximize Conversions I'm not sure it holds anymore. Smart Bidding already uses time of day as an auction signal, so a hard schedule can block conversions the bidder would have bought at a fair price. Say a dental clinic only runs ads from 8 to 6, but a chunk of its form fills come from people searching at 9pm after a toothache. The schedule kills that volume, and the bidder never learns it was there.

I'd still use schedules when the business can't actually take the lead, like a call-only line that's closed, or when the budget runs dry by noon and gets spent on the weakest hours. Before restricting anything, compare conversion rate by hour. If the late-night rate is near zero, cut it. If it's just expensive, let the bidder handle it.

I'm not sure the dental example holds. A 9pm searcher is only worth bidding on if someone can actually pick up that lead. If the form goes to an inbox nobody checks until 8am, the person has probably already booked with a clinic that answered, and you're paying for a conversion that goes nowhere. So I'd schedule by response capacity rather than conversion rate. If the business can't reach a new lead within a couple of hours in a given window, cut that window, even if the late conversions look fine on paper. If someone can call back quickly, leave the hours open and let the bidder price them.

Also, be careful reading hour-level data on small accounts. A clinic might only get a handful of conversions a week, so a "near zero" rate at 9pm can just be noise. I'd wait several weeks before deciding anything.

Response capacity is the part most people skip, but I don't think a hard cut is the right fix. It stops the bidder from learning anything about those hours, and it's a blunt tool for what is really a lead quality problem. If the clinic can import booked appointments or qualified leads back as conversions, Smart Bidding learns which late searches turn into patients and bids on them accordingly. Microsoft Ads supports offline conversion imports for this, so the schedule can stay open and the bidder does the pruning.

Are the 9pm leads being counted as form fills right now, or is anyone feeding back which ones actually booked?

It comes down to whether the booking system can match each lead back to its click. A GCLID, or an enhanced conversion match from hashed email or phone, is what lets the bidder tie a 9pm search to a booked patient. If the clinic only has names in a spreadsheet, most rows won't match, and the bidder ends up learning from very little. Lag matters too: if a booking takes two weeks to close, the import teaches the bidder about those late hours slowly. So check the match rate before deciding the schedule can stay open.

Match rate is mostly decided on the website, before the booking system ever sees anything. The gclid has to be captured in a hidden field on the form and carried through to the CRM, and it quietly disappears if the form lives on a different domain from the landing page or a redirect strips the parameter. Submit a test lead from a real ad click and confirm the gclid shows up in the record before you rely on any import. Enhanced conversions is a separate fallback, and it only helps if the form actually collected an email or phone, which a lot of multi-step dental forms don't do on the first screen.