Hyros tracking is a four-component architecture - client-side tag, server-side integrations, identity resolution, and CAPI pushback - that stitches multi-session, multi-device buying journeys and reports cleaner conversions to ad platforms.
Hyros tracks users across devices and sessions by combining email as a primary identity key with IP and device fingerprint as fallbacks, then pushes stitched conversion events back to Meta, Google, TikTok, and YouTube via each platform's Conversions API.
How Hyros tracking differs from a standard pixel
A standard ad-platform pixel (Meta pixel, Google tag) tracks within a single device and a single browser, relying on cookies that get cleared, blocked, or expire. When a user clicks an ad on their phone, comes back two weeks later on their laptop, and converts, the pixel typically can't connect the journey - the ad-platform algorithm sees an 'organic' conversion and starves the original ad of budget.
Hyros tracking is architecturally different. Instead of relying on cookies alone, Hyros uses email as a primary identity key. When a user gives you their email anywhere in the funnel (opt-in form, checkout, calendar booking), Hyros stitches all sessions associated with that email into a single user journey - regardless of device, regardless of cookie state.
This sounds simple. The implementation is not. Email isn't always available, isn't always consistent across devices (people use different emails for personal vs work), and isn't always captured early enough in the journey to backfill the ad-click attribution. Hyros's identity resolution layer handles the edge cases - IP-based stitching when email isn't yet known, device fingerprint as a secondary signal, and a probabilistic confidence layer that decides whether two sessions are likely the same user.
The output: a multi-touch journey for each known user, with each ad click, page view, opt-in, call booked, and purchase mapped to a single identity. The journey then gets pushed back to ad platforms via their Conversions API as a deduplicated conversion event, giving the platform's optimization algorithm a cleaner signal to bid against.
Signature features
What stands out
Email-keyed identity resolution
The primary stitching mechanism. When a user gives an email anywhere in the funnel, all subsequent sessions associated with that email get merged into a single identity. The most reliable signal Hyros uses.
IP-based fallback stitching
When email isn't yet known, Hyros uses IP address (combined with timestamp proximity and browser characteristics) to make probabilistic stitching decisions. Lower confidence than email-keyed but useful for the pre-opt-in window.
Device fingerprint as secondary signal
Hyros captures a device fingerprint (screen size, user agent, language, timezone, fonts) and uses it as a corroborating signal for cross-session stitching. Doesn't override email-keyed matches; reinforces or contradicts IP-based probabilistic matches.
Server-side event ingestion
Critical conversion events (purchase, call booked, lead) flow into Hyros via server-side integrations from your CRM, payment processor, and calendar tool - not via the client-side tag. Server-side events survive ad blockers, cookie restrictions, and iOS 14's tracking limits.
Conversions API push to ad platforms
Once a journey is stitched and a conversion has occurred, Hyros pushes the deduplicated event back to Meta, Google, TikTok, and YouTube via each platform's Conversions API. The ad-platform optimizer ingests the cleaner signal and rebalances bidding.
Multi-touch attribution model layer
On top of the stitched journey, Hyros applies an attribution model (default: a Hyros blend that weights first-click higher than standard models). Configurable to last-click, first-click, linear, U-shape, or the default blend.
Pricing snapshot
Plans at a glance
Starter
From ~$199/mo
Sub-$1M/yr advertisers
Growth
From ~$499/mo
$1M-5M/yr brands
Scale
Quote-based
$5M+/yr or agency tier
Shuttergen
Tracking shows what won. Shuttergen ships what's next.
Hyros's tracking tells you which ad converted across the multi-device journey; Shuttergen turns that signal into the next 10 variants of your winner. The closed loop from data to creative.
Fit
Who this is - and isn't - for
Best for
- · Funnels where users convert across multiple devices and sessions
- · Sales journeys spanning 30+ days where cookie-based attribution decays
- · Stacks with strong email capture early in the funnel (opt-in, calendar booking, checkout)
- · Brands willing to invest in server-side integration setup for high-quality event data
Skip if
- · Single-session impulse purchases where journey-stitching adds no value
- · Funnels where email capture is rare or late - reduces stitching effectiveness
- · Categories where users actively clear cookies and avoid tracking (privacy-conscious buyer segments)
- · Anyone expecting 100% attribution recovery - even Hyros has gaps
The four components of Hyros tracking, in detail
Component 1: client-side JavaScript tag. A single JS snippet you drop on every page of your funnel. The tag does three things: sets a Hyros first-party cookie containing a unique session ID; emits pageview events and custom events (opt-in form submission, button clicks) to Hyros's backend; reads URL parameters at landing to capture ad-click attribution at the moment of the click. The tag is lightweight (~30kb gzipped) and asynchronous - doesn't block page render.
Component 2: server-side integrations. One integration per CRM, payment processor, and calendar tool. The integrations push conversion events (purchase, call booked, lead submitted) into Hyros's backend via webhooks. Critically: these events flow server-to-server, not through the user's browser. They survive ad blockers, cookie restrictions, and iOS 14's tracking limits. The integration list is heavy on info-product / coaching tools (ClickFunnels, Kajabi, Calendly, GoHighLevel, Skool) and broader ecommerce tools (Stripe, Shopify, ThriveCart).
Component 3: identity resolution layer. Where Hyros earns its keep. The IR layer ingests events from components 1 and 2 and stitches them into user identities using a hierarchy of signals: email (highest confidence) > device fingerprint + IP + temporal proximity (medium confidence) > IP + temporal proximity alone (lowest confidence). The output is a unified journey per known user, with each session, page view, and conversion mapped to a single identity record.
Component 4: outbound CAPI pipeline. Once a journey is stitched and a conversion has occurred, Hyros pushes the deduplicated conversion event back to each connected ad platform via their Conversions API. Meta CAPI, Google Conversion API, TikTok Events API, YouTube Brand Lift API. Each platform's optimizer ingests the cleaner signal and uses it to rebalance bidding toward ads that are actually converting (per Hyros's attribution view, not the platform's native view).
The four components work together. The client-side tag captures the ad click and pre-conversion behavior; server-side integrations capture the conversion itself; the identity resolution layer connects them; the CAPI pipeline reports the result back. Removing any one component breaks the chain.
Identity resolution - how Hyros decides two sessions are the same user
The identity resolution layer is the most technically interesting part of Hyros and the part that most marketing copy glosses over. Here's how it actually works in 2026.
Tier 1: explicit email match (highest confidence). If session A and session B both have the same email associated (because the user opted in or purchased in both sessions, or one session is linked via a CRM webhook to an email), the sessions are merged with 100% confidence. This is the cleanest signal and accounts for most of Hyros's cross-session stitching value.
Tier 2: email-derived fingerprint match. If one session has an email and another doesn't, Hyros checks whether the second session's device fingerprint + IP combination matches a previous session that was email-stamped. If yes, the second session is probabilistically assigned to the same user identity. Confidence: high but not 100% - device fingerprints can collide.
Tier 3: IP + temporal proximity match. When no email is available on either session, Hyros falls back to IP-based matching with temporal proximity (same IP, sessions within 24 hours, similar device fingerprint). Confidence: lower - shared IPs (office Wi-Fi, household routers, mobile carriers) can produce false matches.
Tier 4: device fingerprint alone (rare). Used as a last resort when IP isn't reliable (changing carrier IPs, VPN usage). Confidence: lowest - device fingerprints are designed to be reasonably unique but the signal degrades quickly.
The system surfaces stitching confidence in the dashboard so operators can understand which conversions are 'definitely the same user' vs 'probably the same user'. For ad-platform CAPI pushback, Hyros uses Tier 1 and Tier 2 matches by default; Tier 3 and Tier 4 matches are visible in reporting but not pushed back (to avoid polluting the ad-platform optimizer with low-confidence signal).
Tracking shows what won. Shuttergen ships what's next. Hyros's tracking tells you which ad converted across the multi-device journey; Shuttergen turns that signal into the next 10 variants of your winner. The closed loop from data to creative.
Cross-device stitching - the canonical use case
The killer use case for Hyros tracking is cross-device journey stitching, and the canonical example illustrates why standard pixel tracking fails it.
Journey: a $2,000 coaching offer. The user sees a Meta ad on their phone at 9am Tuesday. They click through, browse the landing page for 3 minutes, opt in to a free training with their email. Wednesday at noon they get a follow-up email, click through on their laptop, watch the 60-minute training. Thursday they book a sales call via Calendly. Friday they attend the call on Zoom. The following Tuesday they convert via Stripe checkout on their laptop.
Standard Meta pixel view of this journey. The Tuesday phone click is recorded. The Wednesday laptop session is unattributed (different device). The Thursday call booking is invisible to Meta (Calendly doesn't push to Meta pixel). The Friday Zoom call is invisible. The Tuesday-following Stripe checkout is recorded, but - because it's a different device than the original click and the click attribution window has expired - it's recorded as an 'organic' conversion with no ad credit. Net: Meta thinks the ad didn't convert and starves it of budget.
Hyros view of the same journey. The Tuesday phone click is recorded by the JS tag, with email captured on opt-in stitching it into the user identity. The Wednesday laptop session is stitched to the same identity via email. The Thursday call booking flows in via the Calendly server-side integration, attributed to the user identity. Friday Zoom call attendance flows in via the Zoom integration. Tuesday Stripe checkout flows in via the Stripe integration. All five touchpoints are stitched into a single multi-touch journey. The conversion is pushed back to Meta via CAPI with attribution to the original Tuesday phone click. Net: Meta sees the conversion, credits the original ad, and increases bidding on similar audiences.
The difference between the two views is the difference between a profitable campaign that gets scaled and an unprofitable-looking campaign that gets killed. For info-product / coaching / high-AOV-DTC brands with multi-device journeys like this, Hyros's tracking architecture is the entire product value.
Where Hyros tracking fails or degrades
Hyros isn't magical. The tracking has well-known failure modes that operators should understand.
Failure mode 1: user never gives email. If a user clicks an ad, browses the landing page, and converts via a non-email path (in-store purchase, anonymous checkout, third-party platform), Hyros has no email key to stitch with. Identity resolution falls back to lower-confidence tiers, and many conversions don't get attributed.
Failure mode 2: user clears cookies between sessions. The first-party Hyros cookie can be cleared by the user, the browser, or by tracking-protection extensions. When the cookie is gone, the next session looks like a new user to the client-side tag. Server-side events still work, but client-side behavior data is reset.
Failure mode 3: user uses different emails per device. Personal email on phone, work email on laptop. Hyros sees two distinct identities and can't merge them without an explicit cross-link signal. This is more common in B2B contexts than DTC.
Failure mode 4: high-volume shared-IP environments. Office Wi-Fi networks, family households, university campuses, mobile carrier NAT - shared IPs make Tier 3 (IP + temporal proximity) matching unreliable. Hyros's confidence layer flags these as low-confidence; the matches don't make it into CAPI pushback but they pollute the dashboard reports.
Failure mode 5: ad platforms tighten CAPI requirements. Meta has periodically tightened CAPI signal-quality requirements (event matching scores, parameter completeness). When platforms tighten, Hyros's pushback can temporarily lose effectiveness while the team updates the implementation. Plan for occasional CAPI burn-in periods.
Internal: see hyros for the product deep dive, hyros-chrome-extension for the browser-extension surface, and hyros-api for the API integration playbook.
FAQ
Frequently asked
How does Hyros tracking work?
What identity signals does Hyros use to track users?
Does Hyros track users across devices?
Does Hyros work after iOS 14?
Is Hyros tracking GDPR-compliant?
What's the difference between Hyros tracking and Meta CAPI tracking?
How accurate is Hyros tracking?
Related
Keep reading
Sources
Tracking shows what won. Shuttergen ships what's next.
Hyros's tracking tells you which ad converted across the multi-device journey; Shuttergen turns that signal into the next 10 variants of your winner. The closed loop from data to creative.