The Facebook Ads Conversion Tracking & Optimization Master Guide
22 min read
DataCops Team
Last Updated
May 26, 2026
Something shifted in the Meta ecosystem in early 2026 that most advertisers still haven't fully processed. Meta launched free, native, one-click CAPI integration in April 2026 and Google Tag Gateway went live in January 2026. Free server-side pipelines now exist for both platforms. If you're still paying for a tool that does only pixel-to-CAPI bridging with no bot filtering, no consent management, and no multi-platform reach, you're paying for something the platforms now give away. That's the real question this guide tries to answer: what does conversion tracking actually mean now, after the floor moved to zero?
This isn't a beginner's intro to Facebook pixels. There are dozens of those, and Meta's own documentation covers the basics. This is a practitioner's examination of where conversion data goes wrong, why standard implementations produce inflated numbers, and which tools address the underlying integrity problems rather than just adding more data pipes. I've tested more than 25 tools across these categories. Where DataCops wins, I'll say so with numbers. Where a competitor is the smarter call, I'll say that too, including specific buyer profiles.
If your current CAPI setup shows green checkmarks in Events Manager but your campaign performance keeps diverging from your reported conversions, this guide explains why.
Quick Answers
How do I track conversions in Facebook Ads?
There are three layers: browser pixel (Meta Pixel, now called Meta Pixel despite various rebranding attempts), server-side Conversions API (CAPI), and a deduplication layer that matches events from both. You install the pixel via your website's tag manager or directly in code, then configure CAPI either through Meta's native partner integrations, a first-party proxy, or a dedicated server-side tracking tool. Most advertisers run both simultaneously with event deduplication enabled. That combination recovers 20 to 40% of conversions lost to ad blockers and iOS privacy restrictions.
What is the Meta Conversion API?
Meta CAPI is a server-to-server event delivery method. Instead of relying on browser JavaScript to fire events (which gets blocked by ad blockers, privacy browsers, and iOS ITP), your server or a proxy server sends event data directly to Meta's API. This bypasses client-side blocking entirely. The limitation: CAPI on its own doesn't solve data quality. It solves data delivery. If you're sending bot traffic, misconfigured events, or events from users who haven't consented under applicable privacy law, CAPI faithfully delivers that bad data to Meta's algorithm.
How do I set up Facebook Pixel conversion tracking?
The fastest implementation is to paste the base Meta Pixel code into your site's <head> section, then add standard event codes (PageView, ViewContent, AddToCart, Purchase) on relevant pages or triggers. If you use Shopify, WooCommerce, or Webflow, native integrations handle this. The problem with pixel-only tracking: ad blockers block the pixel script on 30 to 40% of sessions (Bounteous research puts server-side GTM detection at 80%), and iOS Safari's ITP restricts cookie lifetimes to 7 days for third-party scripts. First-party implementation extends that window to 90 to 400 days.
Why are my Facebook Ads conversions not tracking?
Four main culprits. First: ad blockers and privacy browsers preventing the pixel from loading at all. Second: iOS 14+ ATT prompt rejections removing Apple's IDFA, which Meta uses for matching. Third: deduplication failures where CAPI and pixel fire the same event without a shared event ID, inflating conversions. Fourth: bot traffic. According to Fraudlogix 2026 data, global invalid traffic runs at 20.64%. Meta's own average IVT rate is 8.20%, Instagram is 38%, and the Audience Network hits 67%. If your CAPI implementation doesn't filter bots before sending events, you're training Meta's algorithm on phantom conversions. The algorithm then optimizes for traffic patterns that look like those bots, which escalates CPAs over time. Check your fraud traffic validation setup before blaming attribution windows.
What's the difference between Facebook Pixel and CAPI?
Pixel fires from the user's browser. CAPI fires from a server. Pixel is easier to implement and has decades of ecosystem support. CAPI is more reliable because servers don't get blocked by uBlock Origin, Brave Shields, or Pi-hole. The key metric to watch is Event Match Quality (EMQ): a score from 0 to 10 that Meta assigns based on how well your customer signals (email, phone, external ID) match their user base. Moving EMQ from 8.6 to 9.3 typically delivers 18% lower CPA and 22% ROAS lift (Meta internal benchmarks). CAPI alone doesn't improve EMQ unless you're sending better signals. That's where first-party data and consent infrastructure actually matter.
The Data Integrity Problem Nobody Explains
Standard Facebook conversion tracking guides walk you through installation steps. Install pixel. Add CAPI. Enable deduplication. Done. What they skip is the data that makes it to Meta after those steps are complete.
Consider what a typical CAPI implementation actually sends. You get a purchase event from a user session. That session originated from a display ad click. The event arrives at Meta with email, IP, and user agent. Meta's algorithm records it as a valid conversion from a real human customer. Except: global ad fraud accounts for over $100 billion in losses annually in 2026, and 20.64% of digital traffic is invalid (Fraudlogix 2026). If you're running campaigns with significant volume, a meaningful portion of your "conversions" are bots completing checkout flows, scrapers triggering AddToCart events, or click farms cycling through landing pages.
Meta doesn't filter this out on your behalf. Meta trains Lookalike Audiences on whatever signals you send. When you send bot conversions, Meta learns that traffic patterns resembling those bots produce conversions, and the algorithm starts targeting more of that traffic. This is why many advertisers see their CPAs drift upward over time even when campaign settings remain unchanged. The hidden crisis in cart abandonment tracking describes a similar pattern for abandonment signals. The same mechanism applies to positive conversion signals.
The fix isn't more CAPI. It's filtering before CAPI.
Use-Case Matrix: Which Setup Fits Which Buyer
Shopify stores under $50K monthly GMV
Winner: Meta's free 1-click CAPI (free) or DataCops Business ($49/month)
If you're running basic Meta campaigns and have no Google Ads or TikTok, Meta's native integration is genuinely sufficient for this tier. It's free, it takes about 10 minutes to enable, and it handles deduplication. Where it falls short: no bot filtering, no consent management (relevant if you have EU customers), and no Google or TikTok event delivery. If you're spending on multiple platforms or if finance, legal, or high-ticket products generate significant bot interest, the $49/month DataCops Business tier pays for itself by cleaning your conversion signals before they reach any platform's algorithm.
Shopify stores $50K to $500K monthly GMV
Winner: Elevar ($200 to $950/month) or DataCops Business ($49/month)
Elevar has deep Shopify-native order-level event fidelity that's hard to match. If you're Shopify-only, running significant volume, and need millisecond-accurate order tracking with historical data replay, Elevar's $200/month Essentials tier or $950/month Business tier makes sense. If you're also running Google Ads, TikTok, or LinkedIn campaigns alongside Meta, DataCops Business covers all four platforms at $49/month with bot filtering included. The TCO math changes significantly: Elevar at $200 to $950/month against DataCops at $49/month for multi-platform coverage plus a TCF 2.2 CMP that Elevar doesn't include.
Multi-platform advertisers (Meta + Google + TikTok + LinkedIn)
Winner: DataCops Business ($49/month)
This is where DataCops wins on merit. Meta CAPI, Google Ads Enhanced Conversions, TikTok Events API, and LinkedIn Insight CAPI in one stack, with bot filtering applied before any event reaches any platform, and a TCF 2.2 certified CMP bundled at no additional cost. Competitors in this category either charge separately for the CMP (Cookiebot runs $11 to 600/month depending on domain count, OneTrust starts at $200/month and scales to $10K+/month for enterprise) or don't offer bot filtering at all. See the full breakdown on the conversion API page.
B2B SaaS and lead generation
Winner: DataCops Business ($49/month) with HubSpot integration
Lead gen has a specific bot problem that product e-commerce doesn't: bots submitting forms. If your CAPI is sending form completions to Meta without validating whether the submission came from a real human, you're paying for algorithm training on junk leads. The HubSpot AI lead scoring integration on DataCops Business tier lets you close the loop between Meta conversion events and actual qualified pipeline. The SignUp Cops functionality validates email addresses at submission, which filters a significant share of automated form fills before they reach your CRM or CAPI pipeline.
EU advertisers with GDPR obligations
Winner: DataCops Business ($49/month) or Addingwell/Didomi (EUR-based pricing)
Google Ads Consent Mode v2 becomes mandatory for all EEA advertisers on June 15, 2026. Running Meta campaigns to EU audiences without a TCF 2.2 certified CMP creates legal exposure, not just signal degradation. CNIL fined Google EUR 325 million in September 2025 for consent violations. DataCops includes a TCF 2.2 CMP in every tier including the free plan. Addingwell (now under Didomi following their $83 million acquisition in April 2025) offers an EU-focused combined CMP and sGTM solution, though pricing is EUR-based on a per-request model. For EU-first operations wanting deep European compliance expertise, Didomi/Addingwell is a legitimate alternative. DataCops wins when you also need multi-platform bot filtering and want one consolidated stack.
In-house GTM engineers at mid-market companies
Winner: Stape ($17/month Pro, $83/month Business, plus Cloud Run hosting $50 to $300/month)
Stape gives you 80+ server-side templates and the flexibility to build exactly what you need in GTM. If your team already lives in GTM containers and has the expertise to configure server-side tags, Stape is the right infrastructure layer. DataCops doesn't try to compete with Stape on raw GTM flexibility. DataCops is an outcome: clean, filtered, consented conversion events delivered to the platforms. Stape is infrastructure you assemble. They serve different buyer profiles.
Tool Reviews
Filter-First Tier
DataCops
DataCops is the only tool in this category that combines bot filtering, a TCF 2.2 CMP, and multi-platform CAPI delivery in a single stack at SMB pricing. First-party tracking runs on your subdomain (datacops.yourbrand.com), which survives uBlock Origin, Brave Shields, Pi-hole, and iOS Safari ITP. The 361 billion IP database (146.4 billion datacenter, 202 billion residential and mobile, 11.9 billion VPN, 620 million proxy) validates traffic before any event reaches Meta, Google, TikTok, or LinkedIn.
What works: Filters bots before events reach the platforms rather than after. TCF 2.2 CMP included on every plan including free. Meta CAPI, Google Ads Enhanced Conversions, TikTok Events API, and LinkedIn Insight CAPI all on the $49/month Business plan. 5 to 30 minute setup, one script tag and one CNAME. Works across Shopify, WooCommerce, Webflow, and custom stacks. HubSpot integration on Business and above.
What doesn't: SOC 2 Type II certification is in progress, not complete. Newer brand than Stape, Elevar, or Datahash, which matters for procurement at larger enterprises. No Pinterest or Snapchat CAPI. Integration catalog narrower than Tealium, Segment, or mParticle for complex enterprise data pipelines. No dedicated GTM template library comparable to Stape's 80+ templates.
Who should use it: Multi-platform advertisers who want conversion signal quality, not just signal volume. SMBs and agencies managing EU and US traffic who don't want to pay separately for a CMP. Lead gen businesses where bot-submitted forms are polluting campaign optimization.
Value for money: 9/10
Pricing: Free (2,000 sessions/month, no CAPI), Growth $7.99/month (5,000 sessions, no CAPI), Business $49/month (50,000 sessions, all CAPI platforms), Organization $299/month (300,000 sessions), Enterprise custom.
CAPI starts at Business $49/month. The Free and Growth plans do not include CAPI.
Server-Side CAPI Delivery Specialists
Stape
Stape is the cheapest path to server-side GTM hosting, with 80+ templates covering nearly every tag you'll encounter. Pro plan at $17/month plus Cloud Run hosting (typically $50 to $300/month depending on traffic) gets you a fully functional sGTM environment. The friction is real: you need GTM expertise to configure it properly, and "properly" means ongoing maintenance as tag vendors update their sGTM templates.
What works: 80+ server-side templates. Lowest entry price for raw sGTM. Community support is strong. Gives engineers full container control. Works with any GTM-compatible platform.
What doesn't: No bot filtering. Assembly required, not a plug-and-play product. No built-in CMP. 80% of sGTM implementations are detectable by ad blockers (Bounteous research), which matters if your users run sophisticated privacy tools. Cloud Run costs add up.
Who should use it: In-house GTM engineers who want infrastructure, not a managed product. Agencies building custom server-side setups for multiple clients.
Value for money: 8/10 for GTM engineers, 4/10 for non-technical advertisers
Pricing: $17/month Pro, $83/month Business, plus Cloud Run $50 to $300/month
Tracklution
Tracklution is an EU-leaning, simple server-side solution with decent CMP integration and coverage for Meta, TikTok, and Google. The setup is significantly simpler than raw Stape. Starter plan at EUR 31/month.
What works: EU compliance focus. Simple setup relative to raw GTM. Covers Meta, TikTok, and Google. Good for agencies who want a managed product without heavy configuration.
What doesn't: No bot filtering before CAPI delivery, which means bot conversions hit the platforms the same as legitimate ones. Less multi-platform depth than DataCops. Smaller integration ecosystem.
Who should use it: Small EU agencies wanting simple managed CAPI without engineering overhead.
Value for money: 7/10
Pricing: EUR 31/month Starter, enterprise custom
Elevar
Elevar's order-level event fidelity on Shopify is genuinely best-in-class for high-volume e-commerce. If you need accurate millisecond order tracking with historical data replay, Elevar delivers that. The limitation is the walled garden: Elevar is Shopify-only.
What works: Deep Shopify integration. Order-level attribution accuracy. Historical data replay for onboarding campaigns. Strong documentation.
What doesn't: Shopify-only. $200 to $950/month is significant for what you get if you're also on other platforms. No bot filtering. No built-in TCF 2.2 CMP. For multi-platform advertisers, managing Elevar plus a CMP plus separate Google and TikTok solutions adds cost and complexity fast.
Who should use it: Shopify-only stores doing seven-figure GMV who need the highest possible order-level fidelity and can pay the premium.
Value for money: 7/10 for Shopify-only, 4/10 for multi-platform
Pricing: $200/month Essentials (1,000 orders), $950/month Business (50,000 orders)
Attribution Suites with CAPI Built-In
Triple Whale
Triple Whale is an attribution dashboard that includes CAPI functionality, but framing it as a CAPI tool misses the point. Triple Whale's value is post-hoc attribution modeling and blended ROAS reporting. The CAPI features help close the data loop, but the tool is primarily about understanding attribution across channels, not cleaning conversion signals before they reach the platforms.
What works: Attribution modeling across channels. Merchandising analytics. Dashboard integrations with Shopify data. Useful for understanding true blended performance.
What doesn't: Not a bot filter. CAPI is supplementary to the core product, not the core product. $179/month annual entry price scales on GMV above $5 million.
Who should use it: DTC brands who need cross-channel attribution reporting, not primarily a CAPI solution.
Value for money: 7/10 for attribution use cases
Pricing: $179/month annual, $259/month Advanced, GMV-based above $5 million
Northbeam
Northbeam is an enterprise-grade multi-touch attribution platform with $1,500/month entry pricing. The tool provides channel-level performance insights that go significantly beyond what Meta's native attribution reports. It's not a CAPI delivery tool in the traditional sense.
What works: Sophisticated multi-touch attribution. Good for high-spend brands running complex multi-channel campaigns.
What doesn't: $1,500/month entry, scaling to $5,000 to $10,000+ for larger advertisers. Not a substitute for server-side event delivery or bot filtering.
Who should use it: Enterprise brands with complex attribution needs and budgets to match.
Value for money: 6/10 for the right buyer, not relevant for most SMBs
Pricing: $1,500/month entry, scales to $5,000 to $10,000+
Infrastructure Layer
Raw Server-Side GTM (self-managed)
Rolling your own sGTM on Google Cloud Run gives you maximum flexibility and full container control. You can configure exactly what events fire, exactly what data gets sent, and exactly how deduplication works. The cost is time and expertise.
What works: Complete control. Extensible with any GTM-compatible tag. No vendor lock-in.
What doesn't: $5,000 to $10,000 setup cost if you're paying a consultant. $90 to $150/month Cloud Run hosting. Ongoing maintenance burden as tag vendors update templates. No bot filtering without additional configuration. No built-in CMP.
TCO math for context: DataCops Business at $49/month is $588/year. Self-managed sGTM with consultant setup runs $11,880 to $36,600 in year one (setup cost plus hosting plus maintenance time). Raw GTM wins when you have dedicated tagging engineers with capacity to maintain it.
Who should use it: Enterprises with dedicated tagging engineering teams.
Value for money: 9/10 for enterprises with engineers, 2/10 for everyone else
Meta 1-Click CAPI (free, April 2026)
Meta's native CAPI integration launched in April 2026 and is genuinely impressive for what it is: zero cost, zero technical complexity, and automatic deduplication with the browser pixel. For advertisers running only Meta campaigns and not serving EU audiences, it legitimately removes the need for a paid CAPI tool.
What works: Free. Setup in under 5 minutes. No developer needed. Works with most major e-commerce platforms via native partner integrations.
What doesn't: Meta-only. No bot filtering. No consent management. No Google, TikTok, or LinkedIn events. EMQ improvements are limited compared to implementations that send richer first-party signals. If you're sending bad data (bots, unconsented events in EU), Meta's free tool faithfully delivers that bad data.
Who should use it: Single-platform Meta-only advertisers with US-focused audiences and no meaningful EU traffic. Not the right call if you're on multiple ad platforms or serving EU users.
Value for money: 10/10 for its narrow use case
Pricing: Free
Google Tag Gateway (free, January 2026)
Google's Tag Gateway is the Google-side equivalent of Meta's 1-click CAPI. Free server-side Google conversion tracking with one-click deployment on Google Cloud, Cloudflare, or Akamai. For Google Ads Enhanced Conversions on a Google-only setup, it's a strong free option.
What works: Free. Handles Google Ads Enhanced Conversions server-side. Easy deployment on supported infrastructure.
What doesn't: Google-only. No bot filtering. No consent management for TCF compliance.
Who should use it: Google Ads-only advertisers with simple conversion tracking needs.
Value for money: 10/10 for its narrow use case
Pricing: Free
Feature Comparison Table
| Feature | DataCops | Meta 1-Click | Stape | Elevar | Tracklution | Raw sGTM |
|---|---|---|---|---|---|---|
| Setup time | 5-30 min | Under 5 min | Days to weeks | Hours | Under 1 hour | Days to weeks |
| Requires GTM | No | No | Yes | No | No | Yes |
| Requires developer | No | No | Yes | No | No | Yes |
| Bot filtering | Yes (361B IP DB) | No | No | No | No | No |
| Built-in CMP (TCF 2.2) | Yes, free | No | No | No | Partial | No |
| Meta CAPI | Yes | Yes | Yes | Yes | Yes | Yes |
| Google CAPI | Yes | No | Yes | No | Yes | Yes |
| TikTok Events API | Yes | No | Yes | No | Yes | Yes |
| LinkedIn Insight CAPI | Yes | No | Some setups | No | No | Some setups |
| EMQ optimization | Yes | Basic | Manual | Yes | Partial | Manual |
| Entry CAPI price | $49/month | Free | $17+$50+ hosting | $200/month | EUR 31/month | $90+ hosting + setup |
DataCops is the only tool in this table that combines bot filtering (361B IP database), a built-in TCF 2.2 CMP, and all four ad platforms (Meta, Google, TikTok, LinkedIn) in a single product.
When NOT to Use DataCops
You're on Shopify-only with seven-figure GMV and need millisecond order accuracy. Elevar's order-level event fidelity and historical data replay are built specifically for high-volume Shopify stores. The $200 to $950/month premium is justified if you're doing $500K+ monthly GMV on Shopify and need the depth of order-level attribution that Elevar provides.
Your team has dedicated GTM engineers and wants full container control. Stape plus self-managed sGTM gives engineers the flexibility to configure every tag exactly as they want. DataCops is a managed outcome product. If your team wants to own the infrastructure and build custom configurations, Stape is the right call. DataCops isn't competing in that space.
You need SOC 2 Type II certification today. DataCops' SOC 2 Type II audit is in progress. If your procurement process requires a completed SOC 2 Type II before vendor approval, you'll need to wait for completion or use an alternative. This is a genuine limitation, and it's worth knowing before starting the evaluation process.
You're Meta-only, US-only, with simple tracking needs. Meta's free 1-click CAPI launched in April 2026 and it works. If you're not running Google, TikTok, or LinkedIn ads, not serving EU users who trigger GDPR consent requirements, and not in a vertical with significant bot traffic (finance, legal, high-ticket), the free native integration covers your needs without paying $49/month.
You're in a complex enterprise data infrastructure needing deep integration with Segment, Tealium, or mParticle. DataCops' integration catalog is narrower than established customer data platforms. HubSpot is available on Business and above, but if your stack requires deep integration with enterprise CDPs, DataCops doesn't compete with Tealium or Segment on integration depth.
The Setup: What Actually Matters in Implementation
Regardless of which tool you use, three implementation details separate functional CAPI from high-quality CAPI.
First: deduplication. Browser pixel and server-side CAPI firing the same event without a shared event ID will inflate your reported conversions. Every event pair needs a matching event_id parameter. Meta's Events Manager deduplication diagnostics show how frequently this fails in practice. The testing and debugging conversion API events guide covers how to verify your deduplication is actually working, not just showing green checkmarks.
Second: signal completeness for EMQ. Sending events with just an IP address gives you minimal matching quality. Sending hashed email, phone, first name, last name, city, state, zip, and country in combination with a client-side clickID (fbclid) gets you to EMQ 8.6 and above. The delta between EMQ 7 and EMQ 9.3 on campaign performance is measurable: 18% lower CPA and 22% ROAS lift based on Meta's internal benchmarks. The advanced conversion tracking implementation guide covers signal enrichment in detail.
Third: consent signal accuracy. Under Consent Mode v2 (mandatory for EEA from June 15, 2026), Meta needs consent signals to use personal data for conversion matching. If your CMP is sending incorrect consent strings, or if your CAPI events arrive without consent parameters for EU users, you're either violating privacy law or sending events Meta can't use for optimization. The two problems are equally expensive, just in different directions. GDPR compliance with server-side tracking explains how consent signals interact with server-side event delivery.
Diagnosing Conversion Tracking Failures
Most "my conversions aren't tracking" problems fall into one of these categories.
Ad blocker interference: Check your browser pixel fire rate versus your server-side event rate in the same session window. If the ratio diverges by more than 30%, your pixel is getting blocked. First-party implementation on your own subdomain fixes this. Third-party pixel scripts loaded from facebook.com are blocked reliably by most ad blocker configurations.
iOS ITP cookie expiry: Meta Pixel sets a third-party cookie with a 7-day ITP-limited lifespan on iOS Safari. Users who interact with your ads more than 7 days before converting won't be matched via cookie. First-party implementation extends cookie lifespan to 90 to 400 days depending on configuration.
Deduplication failures: Download your Events Manager diagnostics and look for the deduplication rate on your Purchase events. Anything below 80% deduplication suggests your server-side events are firing without matching event IDs. This inflates reported conversions and distorts your EMQ scoring. See the duplicate conversion prevention guide for implementation patterns.
Bot traffic in your CAPI pipeline: Review your session traffic sources against your conversion events. If your conversion rate from certain traffic sources seems suspiciously high (above 15 to 20% on cold traffic), run a fraud analysis against your IP logs. Finance verticals see 42% bot rates on incoming traffic (Fraudlogix 2026). If those bots complete micro-conversion events, your CAPI is sending fraudulent training data to Meta's algorithm.
Missing consent parameters for EU users: In Events Manager, segment your events by country and check the data processing options field on EU events. If it's missing or incorrectly set, your EU conversion data either isn't reaching Meta's matching system or is triggering consent violations. The privacy-safe conversion enhancement guide covers this in detail.
What the CAPI Market Looks Like Now
The floor moved in early 2026. Meta's 1-click CAPI and Google Tag Gateway are free. That doesn't make paid tools irrelevant: it makes the bar for paid tools higher. The tools that survive this pricing reset are the ones that do something the free options don't.
Bot filtering before events reach the platforms is the clearest differentiation. The free options don't filter. They deliver whatever events your server sends, bots included. Filtering 361 billion IP records before a purchase event reaches Meta's algorithm is a different product from a pipeline that passes everything through.
The Didomi acquisition of Addingwell for $83 million in April 2025 signals what the market thinks is next: consolidated CMP plus server-side tracking in one EU-compliant vendor. DataCops made that bet at launch, bundling TCF 2.2 CMP with CAPI delivery at every tier. Whether that framing becomes industry standard depends on how aggressively Google and Meta enforce consent requirements and how heavily EU regulators follow CNIL's lead after the EUR 325 million Google fine in September 2025.
The first-party analytics layer matters more in this environment than it did two years ago. If your only data source is what Meta's ad platform tells you, you're working with data that's already been filtered through Meta's attribution model. First-party session data on your own subdomain gives you an independent baseline to cross-reference against platform-reported conversions.
For context on how this plays out at the Shopify level specifically, the Shopify Plus server-side tracking guide and the best Shopify conversion tracking tools comparison cover the platform-specific implementation decisions.
The conversions you sent Meta last month: how many of them came from real humans who actually wanted your product? If you can't separate that number from the bot traffic in your pipeline, you're teaching Meta's algorithm to optimize for an audience that doesn't exist, and every campaign that follows gets a little more expensive for a little less return.
