Module 5 · Optimize: Read the Numbers, Make the Calls

The diagnostic tree: which layer is broken?

A red number on the scorecard tells you something is wrong. It doesn’t tell you what. This lesson is the map from symptom to cause — because the single most expensive mistake in week three isn’t a bad ad, it’s fixing the wrong layer.

Think of how you’d handle a slow month in the business. You wouldn’t fire the receptionist because revenue dipped — you’d trace it: fewer inquiries, or fewer inquiries turning into appointments, or fewer appointments turning into engagements? Different break points, different fixes, different people responsible. Your campaign is the same pipeline with the same logic: ad → landing page → lead → booked call → client. Each handoff can fail independently, each failure has its own signature, and the signature tells you where to work.

One habit before the tree: diagnose from the top down, and stop at the first broken handoff. Fixing a landing page is pointless while the ad can’t earn a click; polishing follow-up is pointless while the page produces no leads. Find the earliest failure, fix that, and re-read next week.

The diagnostic tree

1. Symptom: impressions are expensive, or almost nobody clicks. Meta is charging a lot to show your ad, or people see it and scroll past.
   • Broken layer: the creative — the ad itself isn’t earning attention. Remember the auction from Module 1: boring is expensive. When an ad doesn’t get a response, Meta charges you more to force it in front of people.
   • First fix: the hook. Go back to your Module 2 message map and lead with a different door into the same problem — the trigger-moment hook instead of the plain-offer hook, say. Launch it as a new ad alongside the others and let the auction vote. Don’t redesign everything; change the first line and the image, because that’s all anyone saw before scrolling.
2. Symptom: people click, but nobody becomes a lead. You’re paying for traffic; the form stays quiet.
   • Broken layer: the landing page. The ad did its job — a stranger was interested enough to leave their feed. The page lost them. The usual culprits, in order: the page breaks the ad’s promise (different headline, different feel — the “same person mid-sentence” rule from Module 2), the page is slow or clumsy on a phone, or the form asks for too much too soon.
   • First fix: match the scent. Make the landing page headline keep the exact promise of the ad it follows. Then load the page on your own phone, on cellular, and fill the form yourself. Most landing-page failures are found in ninety seconds of pretending to be your own lead.
3. Symptom: leads arrive, but calls don’t get booked. The form fills; the calendar doesn’t.
   • Broken layer: the offer or the follow-up. Meta’s job is finished — it delivered interested strangers. Something between the thank-you page and your calendar is leaking. Two suspects: the follow-up (too slow, too weak, or not happening — check your speed-to-lead standard from Module 4 first, because it fails quietly) or the offer’s bridge (people wanted the free thing but see no natural next step toward talking to you).
   • First fix: audit your own funnel as a lead. Submit the form. How fast did the reply land? Did the thank-you page offer a booking link? Did anyone call? If follow-up checks out, strengthen the bridge — make the booking offer more specific and useful (Module 2’s offer lesson applies to the second ask as much as the first).
4. Symptom: calls get booked, but nobody becomes a client. The campaign fills your calendar; the engagements don’t follow.
   • Broken layer: the sales conversation. Hear this clearly: no ad setting fixes this. The campaign has done everything a campaign can do — it found strangers, earned clicks, produced leads, and put conversations on your calendar. If those conversations don’t convert, the work is in the room, not in Ads Manager.
   • First fix: look at the call itself before blaming lead quality. Are you running the simple call structure from Module 4, or winging it? Are you asking for the engagement at the end, plainly? One honest check: if these same people had come as referrals, would you be closing them? If yes, the leads are fine and the conversation needs structure. If genuinely no — the people are wrong, not just hesitant — then revisit the offer, because the offer selects the audience.

Reading the tree in real life

Take our fictional Meridian Estate Law again, with illustrative numbers. Say a week’s row shows healthy spend, plenty of clicks, and two leads at $100 each — double their $50 allowable. The tree says: clicks but few leads, so it’s the landing-page branch, and the first fix is checking the ad-to-page promise match — not writing new ads, which is where instinct (“the ads must be tired!”) would have sent them. One symptom, one branch, one fix, logged in the decision log, re-read next week. That’s the entire craft.

Watch this step

Frequency and fatigue are easier to diagnose seen — this shows how to read and manage them, one branch of the diagnostic tree this lesson builds. It’s from an independent creator — credited below, so go give them a follow.

9:51