Identity resolution is the process of recognizing that the person browsing on a phone, clicking an ad on a laptop, and completing a purchase on a tablet is the same customer.Documentation Index
Fetch the complete documentation index at: https://docs.upstackdata.com/llms.txt
Use this file to discover all available pages before exploring further.
Why It Matters
Most Shopify visitors browse anonymously. They arrive from an ad, view products, leave, and return days later on a different device. Without identity resolution, each visit looks like a separate person — your attribution data fragments, retargeting audiences shrink, and conversion APIs send events with weak identity signals. The problem intensifies in a privacy-first world. Third-party cookies are disappearing, mobile identifiers are gated behind consent prompts, and cross-domain tracking is increasingly blocked. Merchants need a first-party identity layer that works within these constraints while still connecting the dots across the customer journey. Strong identity resolution also compounds the value of every other part of your data stack. Enrichment becomes more accurate when attached to a unified profile. Conversion API match quality (EMQ) rises when events carry hashed email, phone, and address. Attribution models produce meaningful results only when they can follow a single customer across touchpoints.How It Works
Upstack ID operates at the Resolve stage of the data pipeline, immediately after raw events are captured by Upstack Pixel. The resolution engine uses a layered matching strategy:- Deterministic matching — When a visitor logs in, enters an email at checkout, or clicks a Klaviyo email link, Upstack ID creates a hard link between the anonymous session cookie and the known customer identifier. These matches are exact and high-confidence.
- Session stitching — Events within the same browsing session are grouped using the first-party cookie set by Upstack Pixel. As soon as any event in the session is identified (e.g., checkout), all prior anonymous events in that session are retroactively attributed to the same customer.
- Cross-device linking — When the same email or phone appears across sessions on different devices, Upstack ID merges those profiles into a single identity graph node. Future events on any linked device are automatically resolved.
- Probabilistic signals — Browser fingerprint attributes (viewport, language, timezone, installed fonts) serve as weak tiebreakers when deterministic identifiers are absent. These signals are never used alone — they only reinforce existing deterministic links.
Key Terms
| Term | Definition |
|---|---|
| Identity graph | A data structure that maps all known identifiers (email, phone, cookie, device) to a single customer profile. |
| Deterministic matching | Linking sessions using exact identifiers like email or phone number. |
| Probabilistic matching | Using weaker signals (browser attributes, timing patterns) to suggest — not confirm — an identity link. |
| Session stitching | Retroactively attributing anonymous events to a known customer after identification occurs within the same session. |
| Cross-device linking | Merging profiles when the same identifier appears on different devices or browsers. |
Identity & Stitching Events
See the events Upstack ID emits when profiles are merged or sessions are stitched.
Server-Side Tracking
Learn how Upstack Pixel captures the raw events that feed identity resolution.