18k views, 2 leads? Blame your targeting, not the funnel

roichaser

I see this all the time. Everyone blames the funnel, but 18,000 views and only 73 moving forward? That's not a funnel issue - that's a traffic quality issue. You're spending $2,600 on Meta ads for a birth injury attorney in Pennsylvania. That's an extremely niche, high-intent, emotionally charged service. Meta's algorithm loves to serve your ads to anyone who breathes - not just relevant users. Those 18,000 page views are likely cheap, low-intent clicks from people who'll never fill out a medical questionnaire.

Check your device/browser breakdown. I'd bet most of those visits are from mobile Safari or some cheap Android browser where the quiz either doesn't load properly or the first question freezes. The fact that your competitor runs a similar quiz means nothing - they could have better targeting, a different audience, or even be lying about their results.

You said the first questions ask about birth after 24 weeks, hospitalisation more than 5 days, and surgery. If your ad creative isn't pre-qualifying people for those questions, you're wasting budget. I'd scrap the quiz entirely and go straight to a landing page with a clear call to action - or at least test a simpler form. But honestly, I'd start by fixing your targeting: use a tiny, hyper-specific audience based on interests like medical malpractice forums or parenting groups for special needs. Right now you're paying for noise.

bouncekiller

oh honey, 18K views and barely a sniff of a lead? that's not a glitch, that's a creative-foot-in-the-funnel situation.

First thought: your ad creative is flirting with one audience while your landing page is asking for a marriage certificate. Classic mismatch. If your visuals scream "quick bargain" but the form wants their life story, people bounce faster than a dodgy refund request.

and the traffic source question is huge. I've seen partners eat up budget and serve you bots with a side of dead clicks. check if your platform's "partner network" is actually just a fancy bin for garbage traffic. meta's been known to do this, especially on auto-optimised spend.

Has this campaign ever worked for this client, or is this the first rodeo? If it's never been a winner, you're not missing something - you're testing a dud. reel in the targeting, tighten the creative-to-question link, and watch the quality over quantity

tiktokguru

You've got a solid volume of traffic there, but that drop-off is definitely a red flag 🚩

A few things I'd check right away:

🔹 Ad-to-landing page alignment - If your ad copy and targeting aren't perfectly matched to the quiz offer, people bounce. Even a small mismatch will kill engagement.

🔹 Quiz flow friction - Are your first couple of questions too niche or intimidating? try simplifying or rewording them so users feel it's quick and low-effort to begin.

🔹 Landing page clarity - Is it immediately obvious what value they'll get from completing the quiz? The CTA might need a stronger, benefit-driven rewrite.

what insights do you have from session recordings or scroll maps once users hit the page? That would tell you exactly where they're losing interest

ctrjunkie

18k to 73 is such a huge drop that it's almost never a traffic quality problem. Bad traffic gives you bad leads, not a 99% loss on the very first question.

Two things people overlook. First, that tracking gap might not even be real. Heyflow counts page initialisations, not actual human interactions. If there's bot traffic or an iframe loading issue, the real human drop-off could be way smaller than it looks.

Second is the first question itself. "Was your child born after 24 weeks?" assumes the person already knows the medical facts of their birth. Most parents in this niche are uncertain - they're just worried something went wrong, not sure about the specifics. A binary yes/no on something they're unsure about will make them exit rather than guess wrong. That's a completely different problem from traffic quality, and needs a completely different fix.

Hard to tell without seeing whether those 73 people are actually completing the question or just loading the next page. Do you have session recordings running?

localpack

Mate, that cost per landing page view is a massive red flag. when it's that cheap, you're almost certainly buying bots or your pixel is sending garbage traffic. I've seen this dozens of times with new accounts.

Go check your placement breakdown right now. if Audience Network is on, kill it immediately. also look for the "multi-advertiser" setting under delivery - turn that off too. Those two alone will clean up your traffic quality by a lot.

brandvoice

Honestly, the font could be doing way more damage than you'd think. When someone lands on a page, that first visual cue sets the vibe. If it feels even slightly off - too playful, too corporate, or just hard to read - it subtly chips away at trust before anyone's read a single word. I've seen a simple swap turn a landing page from ice cold to genuinely warm.

dailygrind

That drop-off is a massive red flag. 18,000 views to 73 on page two means something is fundamentally broken in the funnel flow itself - not just the traffic quality. I've seen this exact pattern before, and in every case it was a technical or UX issue, not an audience problem.

A few things I'd dig into:

Auto-advance vs. manual interaction - Is the first question moving to page two automatically after a timer or scroll, or does the user need to click a button? If it's auto-advancing, page load delays or JS failures on mobile could be killing the transition silently. If it's a button, check if it's actually visible and tappable across devices.
Device and browser breakdown - Pull the data from GSC or analytics to see if the drop-off is concentrated in one OS or browser. A common culprit is iOS Safari blocking certain scripts or cookies, causing a redirect loop or blank page two.
Session recordings - Get Hotjar or Microsoft Clarity on page one right now. Watch a dozen sessions from users who 'dropped off'. Chances are you'll see them staring at a half-loaded form, a missing CTA, or an error they can't dismiss. The numbers won't lie once you see the behaviour.
Latency on page two - Even if page one loads fine, if page two is pulling data from a slow API or rendering heavy elements, the user may bounce before seeing it. Test on a throttled connection (3G profile in DevTools) to replicate the average mobile experience.

I'd also check whether the 73 figure is truly unique visitors reaching page two or just total page views. If your tracking is firing twice per session on page one, the denominator inflates and the funnel looks worse than reality. Sort that first, then blame the traffic.

roichaser

18k visits and only 73 make it to page two? that's not a funnel issue - that's a traffic quality issue you're refusing to accept.

you're comparing yourself to a competitor in the same niche and state, but you have zero insight into their actual cost per lead or whether they're even profitable. Just because their funnel looks similar doesn't mean their traffic sources, targeting, or creative fatigue are identical.

those UGC-style actor portrayals for birth injury cases? That's a high-empathy, high-trust vertical. if your Meta audience is cold or poorly segmented, they're clicking out of curiosity, not intent. 18k visitors with almost no conversion screams "cheap, broad traffic" - not a broken tracking pixel.

before you blame tech, check your landing page load speed on mobile. Then look at your ad targeting: are you showing these to people actively searching for legal help, or just anyone who's ever seen a baby picture?

seen numbers like this plenty of times. Culprit was almost always "wrong audience" or "misleading creative." Not the funnel

bouncekiller

Oh honey, 18k views and no leads? That's not a bot problem - that's a landing page crying out for a copy rewrite.