How to Improve Meta Ads Lead Quality: From Junk Leads to Closed Deals

Why Most Meta Ads Generate Junk Leads

The root cause is algorithmic. Meta’s Andromeda algorithm optimizes for whatever event you tell it to find. If your campaign objective is “Leads” and your conversion event is a form submission, Meta finds people who fill out forms [1].

According to Dylan Bradley, this creates the “form-filler trap”: the algorithm identifies users with high form-completion probability, not high purchase probability. The result is a dashboard full of cheap leads that never answer the phone [2].

Multi-expert consensus (Bradley, Charley T from Disruptor School, Andrew Kroeze) confirms the mechanism: if you feed junk leads back as conversions, the algorithm learns to find more junk. Every unqualified lead that fires your pixel trains Meta to repeat the mistake [2] [3].

The solution is a systematic approach across five layers: form filtering, field optimization, event hierarchy, data imports, and measurement.

Layer 1: Higher Intent Forms and OTP Verification

The first filter is the form itself. Meta offers two lead form types: “More Volume” and “Higher Intent.” Higher Intent unlocks OTP (One-Time Passcode) SMS verification, which requires leads to confirm their phone number before submission [1].

According to Ryan Pineda (Leadbase, $100K/month lead gen operation), OTP removes 15-20% of garbage traffic. CPL increases approximately 20%, but the trade is worth it: fake numbers, bots, and low-intent browsers get filtered before reaching your CRM [4].

Key implementation details:

• Select “Higher Intent” form type in the lead form settings (not “More Volume”)
• Turn off autofill to force manual entry, ensuring accurate, up-to-date contact data [1]
• Set sharing to “Open” so users can share the ad organically
• Add a post-form action: direct to Messenger, WhatsApp, or calendar booking immediately after submission

For additional filtering, John Perini (LeadCapture.io, 3.2M qualified leads analyzed) recommends point-of-capture enrichment using tools like Trestle to catch bad numbers, prepaid lines, and TCPA litigator risks before the OTP step [5].

Form Setting	Impact on Quality	Impact on CPL
More Volume (default)	Baseline	Baseline
Higher Intent + OTP	+15-20% garbage removed	+20% CPL
Higher Intent + OTP + manual entry	+25-30% garbage removed	+25-30% CPL
+ Point-of-capture enrichment	+35-40% garbage removed	+30-35% CPL

Layer 2: Field Order and Form Optimization

Form structure directly affects both conversion rate and lead quality. Multi-expert consensus (Pineda, Arsh Sanwarwala, John Perini, Jack Hepp) points to a specific field order and progressive disclosure approach [4] [5].

Recommended field order: First Name, Last Name, Email, Phone Number. Separate first and last name to prevent CRM mapping issues. Ask the least friction information first (email, first name), then progressively disclose remaining fields [4].

Form structure rules:

• Keep total fields under 4 when possible
• Use single-column vertical stacking (outperforms multi-column)
• Add micro-copy on high-friction fields: “Enter your information to save your progress” produced a 2.5x conversion lift in John Perini’s analysis of 3.2M leads [5]
• Place forms above the fold on mobile (images should not push forms down)
• Frame the form as a qualifier: “See if you qualify” outperforms “Contact us”

Adding friction for quality: Jack Hepp recommends adding qualifying questions that mirror the first 3-5 questions your sales team always asks (e.g., company size, current tool, budget range). Total volume drops, but quality increases significantly [6].

For lead gen with heavy qualifying requirements (credit score, exact address), move those fields to a landing page instead of the instant form. According to Silvio Perez (AdConversion), multi-step forms on landing pages increase CVR while maintaining data quality: Step 1 captures the lead (Name, Email), Step 2 asks pre-qualifying sales questions [7].

Layer 3: Baker’s Event Hierarchy (Purchase Optimization, Not Leads)

This is the highest-impact change in the entire stack. Baker’s Event Hierarchy replaces the standard Lead optimization with a three-tier system that forces Meta to find buyers, not form-fillers.

The three-tier event hierarchy:

1. Page View (awareness signal)
2. Lead (form submission, tracked but not optimized for)
3. Engaged Lead / Purchase (the actual optimization target)

According to Charley T (Disruptor School, Meta MBA graduate), the implementation is straightforward: switch campaign objective to “Purchases,” create a custom conversion event for the actual transaction (credit card charged, deal closed), and feed it back via CAPI. One account using this approach scaled to $1M/month with lower ad spend [8].

The Engaged Lead variation: Dylan Bradley’s approach fires a custom pixel event when a lead takes a post-submission action indicating genuine intent, such as clicking a link in the first follow-up email. This creates a middle layer between form submission and purchase [2].

Baker’s Event Hierarchy Formula: Optimization Event = Deepest funnel action with 50+ monthly conversions

Why this works: With standard Lead optimization, Meta can find thousands of form-fillers daily. With Purchase optimization, Meta must identify the behavioral patterns of actual buyers. The algorithm receives fewer but dramatically cleaner signals, which compounds over time.

Average Lead Value (ALV) assignment: Assign a dollar value to your qualified lead action and set the campaign to “Purchases” optimization. This forces Meta to calculate expected value per impression, abandoning form-fillers entirely [2].

Layer 4: First-Click CAPI Imports (Offline Conversion Data)

Standard pixel tracking fails for lead gen because platforms over-attribute. According to John Moran (Tier 11), Meta sees all website traffic (direct, organic, email, other paid) and claims credit for conversions it did not drive. With modeled data and view-through attribution, the problem compounds [9].

The proof: Moran imported 100 conversions to Facebook that never originated from Facebook. All 100 appeared under “1-day view” attribution [9].

Baker uses a three-layer CAPI solution based on Tier 11’s framework:

Layer 1: First-click attribution. Only count conversions where the user’s first-ever site visit came from an actual Meta click. No view-through credit, no cross-channel attribution.

Layer 2: Edge tagging. Tag users at the DNS/CDN level before they reach your website. This captures data before consent mode, cookie blockers, and iOS privacy restrictions strip it away.

Layer 3: Filtered CAPI imports. Import only the conversions you choose: new customers, qualified leads, booked calls. Never import unqualified leads.

CAPI Case Study: Beauty Account (Tier 11)

Week	ENCAC	Trend
Before CAPI imports	$26-27	Baseline
Week 1	$24	-8%
Week 2	$20	-23%
Week 3	$21	-19%
Week 4	$17	-35%
Week 5	$12	-54%
Week 6	$6.73	-74%

Target ENCAC was $10-12. The account beat it within six weeks [9].

Custom New-Customer Events

Standard purchase/lead events let Meta target anyone, including returning customers. Even with exclusions and bid cap set to $0, Meta ignores these controls [9].

Moran’s solution: custom new-customer-only events via Data Suite.

Configuration	Total Conversions	New Customers	New Customer %
Standard event	35	16	~46%
+ Exclude existing + bid cap $0	44	20	~45%
Custom new-customer event	50	42	85%

That is an 80% increase in new customer percentage with custom events [9].

Revenue impact: $937K to $2.1M (126% increase). Spending an additional $1M reduced ENCAC by 8%. Spending an additional $500K reduced ENCAC by 37% [9].

Data quality warnings:

• Shopify direct integration is only 44% accurate for attribution. Do not use it as your CAPI source [9]
• If you are missing the first 5 days of a 30-day sales cycle, your imported CAPI data is unreliable [9]
• Use first-click only (not linear attribution) for accurate new vs. returning customer identification [9]
• Standard Google app integration misses 30-40% of conversions; extreme cases show 10x difference (6,000 server vs. 654 standard) [9]

Layer 5: ENCAC as the North Star Metric

In-platform CPA is a vanity metric. Baker uses ENCAC (Effective New Customer Acquisition Cost) as the single source of truth for Meta Ads performance.

ENCAC = Total ad spend (all platforms) / Total new customers (from CRM)

According to John Moran (Tier 11), with a 3% CTR, 97% of people cannot be tracked via clicks. A large portion of Meta’s reported conversions are modeled (iOS opt-outs, ad blockers, cookie deprecation). The CRM is the only reliable source [9].

How to use ENCAC:

• Calculate monthly across all platforms
• Compare against Meta-reported CAC
• If Meta-reported CAC is much lower than ENCAC: Meta is over-attributing (common with view-through)
• If Meta-reported CAC is much higher than ENCAC: other channels are cannibalizing credit (rare)

Scaling decision framework:

• In-platform CPA: $800 to $900 = acceptable variance
• ENCAC: $50 to $55 = acceptable, keep scaling
• If ENCAC rises more than 15-20% while scaling, pause and investigate [9]

According to Manel Gomez ($35M in Meta spend), ROAS is also a vanity metric until you check audience breakdowns (new/engaged/existing customers). ROAS consistency does not exist; the goal is hitting your average target ROAS every 7 days [10].

CPL Benchmarks by Vertical

Your own historical best 30-day period is the primary benchmark. Industry averages provide secondary context [11].

Vertical	Expected CPL	Cost per Booked Call	Close Rate
Local services	$10-30	$50-150	20-30%
B2C coaching/consulting	$5-20	$100-200	10-20%
B2B coaching/consulting	$25-50	$250-1,000	5-15%
B2B SaaS	$30-80	$150-500	5-15%
Info products	$3-15	N/A (direct sale)	1-5%

CPM diagnostic ranges:

CPM Range	Signal	Action
$8-15	Low-competition vertical	Verify targeting quality; low CPM + high CPL = reaching wrong people cheaply
$15-50	Normal for lead gen	Compare to your own historical data
$50-100	Competitive or narrow audience	Check if vertical justifies it; broadening may reduce CPM
$100-300+	Possible shadowban	Investigate domain quality, account restrictions, ad feedback

Manual Approval CAPI: The Final Quality Gate

For maximum lead quality, Andrew Kroeze and Ryan Pineda recommend a manual approval step before firing the conversion pixel [3] [4].

Implementation with GoHighLevel:

1. Leads enter your CRM pipeline
2. Your team manually reviews each application
3. Only approved leads get dragged to a “Send Pixel” column
4. The column triggers the CAPI conversion event

This prevents unqualified leads from ever training Meta’s algorithm. Combined with Baker’s Event Hierarchy, it creates a closed loop: Meta only receives signals from leads your team has verified as qualified [3].

The quality screening shortcut (Leadbase): Call the first batch of generated leads yourself. Send only the best 3 to the client for a “dynamite” first impression. This builds client trust while you optimize the system [4].

The Complete Baker 5-Layer Lead Quality Stack

Layer	Tactic	Impact
1. Form Filtering	Higher Intent + OTP + manual entry	Removes 25-30% garbage
2. Field Optimization	Progressive disclosure + qualifying questions	Increases quality without killing volume
3. Event Hierarchy	Purchase/Engaged Lead optimization	Shifts algorithm from form-fillers to buyers
4. CAPI Imports	First-click, edge tag, filtered imports	Trains algorithm on clean data only
5. ENCAC Tracking	CRM-based measurement	Provides single source of truth

Each layer compounds. Form filtering removes obvious junk. Field optimization qualifies the remainder. Event hierarchy redirects the algorithm. CAPI imports clean the training data. ENCAC tracking ensures you measure what matters.

A Baker client running all five layers reduced ENCAC by 74% in six weeks while increasing new customer percentage from 46% to 85% [9].

---

Sources

1. Ben Heath, Matt Shiver. “Meta Lead Gen Best Practices: Higher Intent Forms, OTP Verification, Autofill Settings.” Meta Advertising Consensus Framework, 2026.
2. Dylan Bradley. “Engaged Lead Pixel Strategy and Purchase Optimization for Lead Gen.” Lead Generation Strategy Framework, 2026.
3. Andrew Kroeze. “Algorithm Training and Manual Approval CAPI for Lead Quality.” YouTube Educational Series, 2026.
4. Ryan Pineda, Leadbase. “OTP Verification Impact, Field Order Optimization, and Lead Screening.” $100K/Month Lead Gen Operation Case Study, 2026.
5. John Perini, LeadCapture.io. “Form Optimization Analysis: 3.2 Million Qualified Leads.” Point-of-Capture Enrichment and Micro-Copy Testing, 2026.
6. Jack Hepp. “Qualifying Questions and Friction-for-Quality in Lead Forms.” Lead Gen Optimization Framework, 2026.
7. Silvio Perez, AdConversion. “Multi-Step Landing Page Forms and High-Intent Source Exhaustion Strategy.” B2B Lead Gen Methodology, 2026.
8. Charley T, Disruptor School (Meta MBA graduate). “Purchase Optimization for Lead Gen and One CBO Structure.” Andromeda-Era Campaign Architecture, 2026.
9. John Moran, Tier 11. “First-Click CAPI Imports, Edge Tagging, and Custom New-Customer Events.” Beauty Account Case Study: ENCAC $26 to $6.73, 2026.
10. Manel Gomez. “ROAS as Vanity Metric, Audience Breakdowns, and Weekly Creative Cadence.” $35M Meta Spend Operational Framework, 2026.
11. Baker Team. “Meta Ads CPL Benchmarks by Vertical and CPM Diagnostic Ranges.” Internal Client Data Analysis, 2026.