Cross-Channel Identity Resolution
What it is
Cross-Channel Identity Resolution is the BotBat CDP's mechanism for recognizing that the same person interacting across multiple channels (WhatsApp, email, SMS, web) is a single customer who should have one unified profile. When a customer messages you on WhatsApp using their phone number, then fills out a web form with their email address, and later clicks an SMS link, identity resolution links these interactions together using shared identity keys such as email, phone number, and external ID.
The result is a single, merged profile with a complete cross-channel history. This unified view enables accurate segmentation, personalized messaging, and a consistent customer experience regardless of which channel the customer uses at any given moment. Without identity resolution, the same person appears as multiple separate profiles, leading to duplicate communications, fragmented analytics, and an inconsistent experience.
Identity resolution operates in real time. Every incoming event is evaluated against configured identity keys, and matches trigger automatic profile merging according to your chosen strategy. For cases that require human judgment, the system provides a merge queue where your team can review and approve suggested merges before they take effect.
When to use
| Scenario | Description |
|---|---|
| Multi-channel engagement | Your customers interact through more than one channel and you need a unified view. Without identity resolution, the same person appears as separate profiles in WhatsApp, email, and web. |
| Onboarding from multiple sources | You import contacts from different systems (CRM, email platform, support desk) that use different identifiers. Identity resolution matches records by shared keys and merges them. |
| De-duplicating contacts | Your People list has grown organically and contains duplicate profiles. Identity resolution rules and manual merge tools clean up the database. |
| Enriching anonymous profiles | An anonymous web visitor later identifies themselves via a WhatsApp message or email opt-in. Identity resolution retroactively links their anonymous browsing events to their known profile. |
| Maintaining accurate audience counts | Duplicates inflate segment sizes and campaign audience counts, leading to inaccurate analytics and wasted message credits. Identity resolution ensures each person is counted once. |
| Configuring matching rules | You need to define which identity keys to use, set matching priorities, and control how conflicts are resolved during merges. |
Steps
Step 1: Navigate to Identity Resolution settings
Click "Customers" in the main sidebar, then go to Settings > Identity Resolution to view and configure the identity matching rules. This page is the central control panel for all identity resolution behavior in your workspace.
Step 2: Review identity keys
BotBat supports the following identity keys by default. Each key type has its own matching logic and normalization rules.
| Identity Key | Matching Logic | Normalization |
|---|---|---|
| Matches profiles sharing the same email address. | Case-insensitive comparison; whitespace trimmed. | |
| Phone number | Matches profiles sharing the same phone number. | Normalized to E.164 format (e.g., +15551234567). |
| External ID | Matches profiles sharing a custom external identifier (CRM ID, application user ID). | Exact string match; no transformation applied. |
| Cookie/Device ID | Matches anonymous web sessions to profiles when the same browser is later identified. | Exact string match on the tracking cookie value. |
Step 3: Configure matching priorities
Set the priority order for identity keys. When an incoming event contains multiple keys that could match different existing profiles, the highest-priority key determines which profile receives the event. For example, if email has higher priority than phone, and an incoming event contains both, the system first attempts to match by email. If that match succeeds, the phone number is added to the matched profile. If not, the system falls back to phone matching.
Step 4: Set merge strategy
Define how property conflicts are resolved when two profiles are merged into one. The platform offers three strategies.
| Strategy | Behavior | Best For |
|---|---|---|
| Most recent wins | The property value from the most recently updated profile is kept. | Fast-moving data where the latest value is most accurate. |
| Primary wins | The property value from the profile designated as "primary" (usually the older or more complete profile) is kept. | Established databases where older records are more reliable. |
| Manual review | Conflicting profiles are flagged for human review instead of being auto-merged. | High-value records where incorrect merges could cause significant problems. |
Step 5: Enable automatic identity resolution
Toggle on automatic matching. When enabled, every incoming event is evaluated against identity keys in real time. If a match is found, the event is attributed to the existing profile. If multiple profiles match, they are merged according to the configured strategy. If no match is found, a new profile is created.
Step 6: Review the merge queue
Navigate to Identity Resolution > Merge Queue to see profiles flagged for manual review. Profiles land in the queue due to conflicts, low confidence matches, or multiple potential matches. Each entry shows the candidate profiles, their overlapping identity keys, and a confidence score.
Step 7: Manually merge two profiles
From any profile detail page, click "Merge" and search for the duplicate profile. The platform displays a side-by-side comparison showing properties, events, and segment memberships from both profiles. For each conflicting property, choose which value to keep. Review the event timelines carefully before confirming, because merged events cannot be easily separated after the fact.
Step 8: View the identity graph
On a profile detail page, click "Identity Graph" to see a visual representation of all identity keys linked to this profile. The graph displays each key (email, phone, device ID) as a node connected to the unified profile, along with timestamps indicating when each key was first and last seen. This visualization helps you understand how the profile was constructed and which channels contributed identity data.
Step 9: Monitor resolution metrics
On the Identity Resolution dashboard, review key metrics that indicate the health of your identity resolution process.
| Metric | What It Tells You |
|---|---|
| Total merges (auto + manual) | The overall volume of profile merges, broken down by automatic and manual. |
| Duplicate detection rate | The percentage of incoming events that matched an existing profile rather than creating a new one. |
| Merge queue size | The number of profiles awaiting manual review. A growing queue indicates the need for more frequent reviews or adjusted matching rules. |
| Profiles with multiple identity keys | The count of profiles linked to more than one identity key, indicating successful cross-channel resolution. |
Common pitfalls
| Pitfall | Why It Matters | How to Avoid It |
|---|---|---|
| Overly aggressive matching | Matching on weak keys (e.g., first name + city) can incorrectly merge two different people. | Use strong, unique identifiers like email and phone as primary keys. |
| No normalization | The same phone number in different formats (+1-555-123-4567 vs. 15551234567) will not match. | Ensure normalization rules are active for all identity key types. |
| Ignoring the merge queue | Profiles flagged for review sit indefinitely if not addressed, degrading data quality. | Schedule regular reviews (daily or weekly depending on volume) to keep the queue manageable. |
| Merging without reviewing events | All events from the secondary profile transfer to the primary. Incorrect merges are difficult to reverse. | Always review event timelines of both profiles before confirming a merge. |
| Relying solely on automatic resolution | Edge cases (shared family email, company phone used by multiple employees) require human judgment. | Monitor false-positive merge rates and adjust matching rules accordingly. |
| Late identity key setup | Starting without identity resolution configured causes duplicates that are harder to clean up later. | Configure identity keys and matching rules before importing data or launching channels. |
Implement a "progressive identification" strategy: start tracking anonymous visitors with a cookie or device ID, then promote them to a known profile as soon as they provide an email or phone number. This links pre-identification browsing behavior to the customer's permanent profile, giving you a complete picture of their journey from first touch.
- Identity Resolution Settings
- Identity Keys
- Merge Queue
- Manual Profile Merge
- Identity Graph
- Resolution Metrics