Identity resolution is the process of matching data points — cookies, device IDs, email addresses, phone numbers — across channels and touchpoints to determine they belong to the same person. The output is a unified customer profile that gives marketers a complete view of the buyer journey, enabling accurate attribution and personalised messaging at scale.
Without identity resolution, the same person browsing your site on mobile, clicking your email on desktop, and converting through a retargeting ad looks like three different customers. That means inflated reach numbers, duplicate suppression failures, and attribution that consistently lies to you.
What is identity resolution?
Identity resolution is a data process that connects fragmented signals about a person — their device identifiers, browser cookies, email hash, phone number, loyalty ID — and resolves them into a single canonical record. That record is sometimes called a "golden record" or unified profile.
There are two matching methods:
- Deterministic matching — uses definitive identifiers like a hashed email or logged-in user ID. High accuracy, lower coverage. When someone logs in on both mobile and desktop, you know with certainty it's the same person.
- Probabilistic matching — uses behavioral signals (IP address, device type, browsing patterns) to infer a match. Higher coverage, lower confidence. Used when you don't have a logged-in signal.
Most enterprise identity resolution systems combine both: deterministic for known users, probabilistic to fill the gaps.
The average customer touches 6.4 different channels before converting (McKinsey, 2024). Without identity resolution, each touchpoint looks like a separate anonymous visitor. With it, you see the full path and can credit the right channels.
Why identity resolution matters for marketers
Fragmented data is expensive. When you can't connect customer touchpoints, you overspend on acquisition, underinvest in retention, and make channel budget decisions based on incomplete attribution. Identity resolution solves all three:
- Accurate attribution. When you know a customer saw your LinkedIn ad, read your blog, opened your email, and then converted from a remarketing ad, you can credit and fund each channel appropriately — not just the last click.
- Reduced wasted spend. A fitness studio implementing Facebook ad tracking with identity resolution principles cut its cost-per-lead by 40% within three months by eliminating duplicate targeting and identifying creatives that actually drove conversions.
- Personalisation at scale. A unified profile shows where someone is in the buyer journey. A prospect who read three product comparison articles should get a different message than someone who just landed from a branded search.
- Suppression accuracy. Exclude existing customers from acquisition campaigns correctly. Without identity resolution, you're paying to advertise to people who already bought.
- Compounding returns. Unlike paid media that stops delivering when the budget runs out, the data infrastructure built through identity resolution keeps improving future campaigns automatically.
How identity resolution works in practice
A working identity resolution system has three layers:
Sources: CRM + website events + email platform + ad platforms + POS
# Layer 2 — Identity matching
Deterministic: hashed email / login ID → exact match
Probabilistic: IP + device + behavior → confidence-scored match
# Layer 3 — Unified profile output
Golden record: customer_id: 00142 | devices: 3 | sessions: 47 | stage: consideration
The system continuously updates profiles as new data arrives. A customer who signed up for your newsletter yesterday gets linked to the anonymous browser sessions they had three months ago — once you have a deterministic identifier to anchor the match.
The role of a Customer Data Platform (CDP)
A CDP is the most common tool for operationalising identity resolution. It ingests data from all connected sources, runs the matching logic, maintains the unified profile store, and sends segmented audiences downstream to ad platforms, email tools, and personalisation engines. Popular CDPs include Segment, Salesforce Data Cloud, Adobe Real-Time CDP, and mParticle.
Types of identity resolution — deterministic vs probabilistic
| Type | Signal used | Accuracy | Coverage | Best for |
|---|---|---|---|---|
| Deterministic | Email hash, login ID, phone | Very high (95%+) | Low-medium (known users only) | Logged-in products, email lists |
| Probabilistic | IP, device, behavior patterns | Medium (70-85%) | High (anonymous users) | Awareness campaigns, prospecting |
| Hybrid | Both combined | High overall | High | Enterprise marketing stacks |
| First-party only | Your own CRM + pixels | High for known customers | Limited to your database | Privacy-first, cookieless contexts |
Real identity resolution examples
1. Local fitness studio — paid media efficiency
A fitness studio ran Facebook ads targeting "people interested in fitness" in their city. Without identity resolution, the same person was being targeted across three device profiles simultaneously. After implementing a CDP that connected Facebook pixel events to email-hashed CRM records, they eliminated duplicate targeting and identified the two creative variants actually driving gym sign-ups. Cost-per-lead dropped 40% in three months.
2. B2B SaaS — full-funnel attribution
A software company couldn't understand why organic content drove so many first-touch sessions but appeared to close almost no deals in their last-click model. Identity resolution connected the anonymous blog reader sessions to CRM contacts — revealing that 68% of closed deals had consumed at least three organic articles before requesting a demo. Budget shifted from paid to content accordingly.
3. Ecommerce — suppression accuracy
An ecommerce brand advertising on Google and Meta discovered through identity matching that 23% of their "prospecting" ad impressions were going to existing customers — because their CRM email list didn't account for customers who used different email addresses at checkout. Fixing the match rate reduced wasted acquisition spend by $14,000/month.
Identity resolution vs customer segmentation — what's the difference?
Both concepts work with customer data, but at different stages of the data pipeline.
Identity resolution
- Happens at the data layer — before analysis
- Answers: "Is this the same person?"
- Produces unified profiles (golden records)
- Input to segmentation
- Requires data engineering or a CDP
Customer segmentation
- Happens at the analysis/activation layer — after unification
- Answers: "Which group does this person belong to?"
- Produces audience segments for targeting
- Output from unified profiles
- Requires marketing strategy, not just data tooling
6 best practices for identity resolution
- Start with first-party data. Your CRM, email lists, and on-site login data are the most reliable deterministic identifiers. Build from there before purchasing third-party identity graphs.
- Standardise your data schema first. Identity matching fails when email addresses are stored differently across systems (uppercase vs lowercase, trailing spaces). Normalise inputs before attempting to match.
- Treat identity resolution as ongoing, not one-time. Customer profiles change — emails change, devices change, consent changes. Build a continuous matching process, not a one-off import.
- Prioritise consent and data compliance. GDPR and CCPA both have explicit requirements around processing identifiable data. Every identity graph needs a lawful basis and a clear consent record.
- Measure match rates as a KPI. Track what percentage of your anonymous sessions you can resolve to a known customer. A match rate below 20% indicates a data gap worth closing.
- Connect identity resolution directly to ad platforms. Matched audiences uploaded to Meta, Google, and LinkedIn outperform cold targeting by 2-4x for conversion rates on average.
High impression counts don't mean your identity resolution is working. You can show an ad to the same person 12 times across 4 devices and count it as 12 unique reaches without ever knowing it was one person. Track resolved profile count and match rate, not just raw reach metrics.
Common identity resolution mistakes to avoid
- Relying solely on third-party cookies — Chrome's third-party cookie deprecation removes the most common probabilistic signal. First-party data is now the foundation.
- Treating it as a one-time data project — customer data constantly changes. Identity graphs degrade without continuous refreshing.
- Skipping consent management — matching identifiers without explicit consent creates regulatory exposure under GDPR and CCPA.
- Over-investing in probabilistic matching at the expense of first-party data collection — improve your login rates and email capture before spending on identity graph vendors.
- Not measuring match rates — if you don't track what percentage of unknown sessions resolve to known profiles, you can't improve the system.
Frequently asked questions
Identity resolution is the process of matching data points — device IDs, cookies, email addresses, phone numbers — to determine that they all belong to the same person. The result is a unified customer profile you can use across marketing channels.
Start with your first-party data: email lists, CRM records, and on-site behavior tracked by your analytics platform. A customer data platform (CDP) connects these sources and begins matching records. Build from there before layering in third-party data.
Yes for businesses with multiple customer touchpoints. Identity resolution reduces wasted ad spend, improves attribution accuracy, and lets you personalise messaging based on full journey context. Businesses that implement it consistently outperform those running siloed campaigns.
Early signals — improved match rates, reduced duplicate profiles — appear within 4-8 weeks. Meaningful downstream impact on conversion rates and reduced cost-per-acquisition typically shows in 3-6 months, depending on data volume and marketing channel mix.
Identity resolution is a process — matching and unifying data. A customer data platform (CDP) is a tool that performs identity resolution automatically. CDPs ingest data from multiple sources, apply deterministic or probabilistic matching, and output unified profiles for downstream use.
Related glossary terms
Sources
- [01]McKinsey — The value of getting personalisation right
- [02]Segment — What is identity resolution?
- [03]IAB — Identity solutions for digital advertising
- [04]Forrester — Identity Resolution Platforms Wave
- [05]Internal client audit: identity resolution implementation reducing CPL 40% — theStacc case study, 2026
