Improve the system from your own data, one variable at a time

Why it matters

After one pilot you have something no benchmark can give you: your own cost per lead, qualification, booking and show rates in your market with your follow-up. Fixing the weakest stage usually moves cost per held appointment more than any change to targeting, and most of those fixes cost nothing.

Step by step

1. 01

   Replace the calculator assumptions with actual rates

   From the tracker totals: actual CPL = spend ÷ CRM leads; qualification = qualified ÷ leads; booking = booked ÷ qualified; show = held ÷ booked. Put them in the calculator (chapter 11). Keep your historical close rate until you have enough held appointments from ads to measure it.

   Check: The calculator shows your pilot's actual rates and a new break-even cost per held appointment.

2. 02

   Break results down by ad

   In Ads Manager, open the Ads tab with the pilot dates selected and note spend, leads and cost per lead per ad. Then, in your CRM, count qualified and held per ad using the form and ad information on each lead. With small numbers, look for big differences (one ad with twice the CPL of the others), not small ones.

   Check: The review sheet lists every ad with spend, CPL and at least qualified counts.

3. 03

   Find the weakest stage

   Biggest drop is at	Likely cause	One change to test
   Spend to leads (CPL high)	Creative or offer does not stop the scroll	Replace the weakest ad with a new angle; keep the form
   Lead to qualified	Offer attracts curiosity, form too easy	Make one question stricter (for example, timeline) or add phone verification
   Qualified to booked	Call script, booking ask, calendar availability	Offer two specific times on the first call; add evening slots
   Booked to held	No reminders, weak confirmation	Add a confirmation call the day before plus a reminder text where consent allows
   Held to signed	Appointment quality	Outside this part: work on your listing or buyer presentation

   Check: You named one stage and the number that shows it is weakest.

4. 04

   Commit to one change and build it as a duplicate

   Write the single change on the review sheet. For an ad or form change, duplicate the ad or form and rename it with a new version (_V4, _v2), rather than editing in place: Meta's help says edits replace the old version and you cannot separate metrics from before and after an edit, and some edits restart the learning phase or send the ad back to review. Keep the new version in draft until you decide to run the next 14 days.

   Check: Exactly one new ad, form or script version exists, in draft, with a new version name.

5. 05

   Check the system before scaling

   Only raise the budget after a 14-day run lands under your ceiling and your responder can handle more leads at the same speed. If speed misses rise when volume rises, the extra spend is wasted. Raise in modest steps and treat each raised budget as a new 14-day pilot with the same rules.

   Check: The review sheet states either "same budget" or a new budget with a reason tied to capacity and cost per held appointment.

Copyable templates

PILOT REVIEW: [campaign name]   Dates: ____ to ____

RESULTS (from tracker, CRM counts)
Spend $____  Leads ____  Qualified ____  Booked ____  Held ____  Signed ____
Cost per lead $____  per qualified $____  per booked $____  per held $____
Ceiling per held $____   Verdict at day 14: continue / iterate / kill

BY AD (Ads Manager, ad level, same dates)
Ad name                         Spend   Leads   CPL    Qualified (CRM tag)   Held
_______________________________ ______  _____   _____  ___________________   ____
_______________________________ ______  _____   _____  ___________________   ____
_______________________________ ______  _____   _____  ___________________   ____

WEAKEST STAGE (largest drop vs my assumption in the calculator)
Stage: lead->qualified / qualified->booked / booked->held / held->signed
Assumed ___%  Actual ___%

ONE CHANGE FOR THE NEXT 14 DAYS
Variable: creative / offer / form question / follow-up script / reminders
Exactly what changes: ____________________________________________
Everything else stays the same: yes
New names: ad ______ form ______ (duplicate, never edit in place)

CALCULATOR UPDATE
Replace assumptions with actuals: CPL ___  qual ___%  book ___%  show ___%
New break-even cost per held $____

Try it

How to test it

Ask: if the next 14 days go better, will I know why? If your review names more than one change, cut it back to one. Then confirm the new version is a duplicate with a new name and is still a draft.

Troubleshooting

Advanced technical details

Optional. You can follow every step above without these.