Quick answer
CIAM, customer identity and access management, covers how customers register, sign in, manage their profile and control how their data is used. It is IAM aimed at external users, designed for low friction at large scale. Analytics then measures what those identified users do.
Formula: Sign-up completion rate = accounts created ÷ registrations started
What is CIAM?
CIAM stands for customer identity and access management: the systems and rules that let people outside your company, mainly customers, register, prove who they are, sign in, manage their profile and control how their data is used. IBM describes it as managing the digital identities of external users, covering profile data, verification and secure access to online services such as e-commerce sites.
The two halves of the name do different jobs. Identity is the record: who this customer is, which email, phone or social account they use, what they have agreed to. Access is the gate: letting the right person in, at the right level, and keeping the wrong one out. A sign-up form, a "sign in with Google" button, a password reset email, a one-time code and a "manage my consent" page are all CIAM.
CIAM is a different thing from the analytics that measures it. It decides who the user is and whether they get in. Analytics counts what that user does after. They meet at one important place, the user ID, which is why a CIAM glossary page belongs next to user ID and identity resolution.
How is CIAM different from IAM?
IAM (identity and access management) in the traditional sense controls how an organization's own employees and contractors reach internal systems. CIAM is the same discipline pointed at customers, and the differences come from who the users are.
| Dimension | IAM (workforce) | CIAM (customers) |
|---|---|---|
| Users | Employees, contractors, partners; thousands | Customers and prospects; up to millions |
| Who creates the account | The company (HR onboarding) | The customer, by self-service sign-up |
| Priority | Security and control | Low friction first, with security built in |
| What a bad experience costs | A support ticket | A lost sale or an abandoned sign-up |
| Typical sign-in | Single sign-on to company apps, enforced MFA | Social login, email, passwordless, MFA as an option or risk-based |
| Data handled | Job role, department | Profile, preferences, consent, purchase and contact data |
| Privacy focus | Internal policy | Consumer privacy law and consent records |
IBM puts the contrast in terms of user experience: customer flows are designed to be smooth because frustrated customers can simply leave, while employees cannot easily walk away from internal systems. Many products now serve both kinds of user, but the design pressures stay different.
What does a CIAM system actually do?
Capabilities listed by IBM, with the event an analytics tool would want from each:
| Capability | What it does | Event worth tracking |
|---|---|---|
| Self-service registration | Lets a customer create an account, often in stages | signup_started, signup_completed, method used |
| Authentication | Password, social login, passwordless, MFA, adaptive checks | login_success, login_failed, mfa_challenge |
| Single sign-on | One login across several of your apps | sso_login, app name |
| Profile and preference management | Customer updates their own details | profile_updated (field names, not values) |
| Consent and settings | Customer controls how data is used; records are kept for audit | consent_updated, categories only |
| Security controls | Fraud and account-takeover detection, identity proofing | account_locked, verification_required |
The right-hand column is the practical bridge. A CIAM system has the full detail; analytics only needs the event name and a few non-sensitive properties. Never send passwords, one-time codes or other secrets into analytics, and keep PII out of event properties unless there is a specific reason.
What does CIAM have to do with analytics?
Three connections, all of them useful:
- CIAM issues the stable ID. Analytics identifies anonymous browsing by device or cookie, which changes by browser and device. The CIAM account ID is the first identifier that stays the same across devices, so it is what you pass to analytics at login. Without it, one customer is several visitors; see data silo for what that does to counts.
- CIAM's consent choices should decide what analytics collects. If a customer declines analytics, the consent record has to reach the measurement layer. A consent management platform is the usual link.
- Analytics measures the CIAM experience itself. Sign-up and login are among the highest-friction steps in a product. The funnel from "started registration" to "first login" is where CIAM design decisions show up as numbers.
Worked example: how login method changes sign-up conversion
Sample data: a store offers two ways to create an account. 10,000 people start registration in a month, 4,000 on the email-and-password form and 6,000 with a social login button.
| Method | Started | Completed | Completion rate |
|---|---|---|---|
| Email and password | 4,000 | 2,480 | 62.0% |
| Social login | 6,000 | 4,860 | 81.0% |
| All | 10,000 | 7,340 | 73.4% |
How to recompute: completion = completed ÷ started (2,480 ÷ 4,000 = 62.0%; 4,860 ÷ 6,000 = 81.0%). The blended rate is (2,480 + 4,860) ÷ 10,000 = 73.4%. The 19-point gap means the password form loses 1,520 people (4,000 − 2,480) against 1,140 on the social button (6,000 − 4,860), even though 2,000 fewer people started it.
What to do with it: if the email form is where people stall, test fewer fields, a visible password rule, or passwordless login, and re-measure. Note the limit: the table shows that the methods differ, not why, and people who choose social login may differ from those who do not. Treat the gap as a lead for a test, not a proven effect. This is the CIAM-versus-analytics split in miniature: CIAM provides the sign-up flow, analytics provides the numbers that tell you whether the flow works.
What are the common mistakes?
- Using email as the permanent ID. Emails change and are personal data. Use an internal account ID as the analytics user ID and treat email as an attribute.
- Sending secrets or raw personal data to analytics. Event names and non-sensitive properties are enough. Hash or omit identifiers where you can.
- Treating consent as a one-time banner. Consent can change. The measurement layer has to follow the current choice.
- Adding friction for security without measuring it. Every extra step costs sign-ups. Measure the funnel before and after a change.
- Mixing up CIAM and a CDP. CIAM manages who the customer is and whether they can log in. A customer data platform combines data from many sources for activation. They overlap on profiles but are different tools.
- Assuming a vendor makes you compliant. A CIAM tool can support consent and auditing, but compliance depends on how you configure and use it and on the law that applies to you. This page is not legal advice.
CIAM and TapCub
TapCub is not a CIAM product. It does not register customers, sign them in, store passwords, run MFA or single sign-on, or manage a customer's account and consent preferences for your application. Your CIAM, or your own login system, does that. TapCub starts after sign-in and measures what logged-in people do.
- Pass the stable ID at login. After your CIAM signs someone in, call
tapcub.login('u_123', { $name: 'Ada' })with your internal account ID. The call can also take an identifier type such as email, phone number, OpenID or UnionID; email is lower-cased and hashed before use, and phone numbers are normalized. Pass an internal ID rather than raw personal data wherever you can. - See the person across devices. In identified mode, which requires the Pro plan or above and is switched on by contacting support, the person shows up in All-platform analytics › Users › People with one timeline across web and app. All-platform analytics › Users › Identity map logs each bind, split and rejected value. Anonymous visitors never get a profile.
- Call
tapcub.logout()when the session ends. The next browsing on that browser is no longer attached to the previous user until the next login. - Match collection to consent. In Configuration › Privacy center, collection tiers and region rules decide what is recorded. Under the default rule for the European Economic Area and the UK, consent is required, and without consent for analytics collection drops to the anonymous tier while page views and custom events are still recorded. TapCub applies this to its own collection. It does not manage the consent record inside your application.
- Measure the sign-up funnel. In All-platform analytics › Behavior › Behavior analysis, build the funnel from
signup_startedtosignup_completedto first login, and split by method, like the example above. Event names and properties are planned in All-platform analytics › Data › Tracking plan.
Sources
- IBM, What is customer identity and access management (CIAM)?.
- NIST, SP 800-63-3 Digital Identity Guidelines (identity proofing, authentication and federation for public-facing services; check for the current revision).
Was this explanation clear?
We use your answers to fix and expand this page.
Couldn’t send. Please try again in a minute.
Thanks for the feedback. We’ll use it to improve this page.You’ve already sent feedback on this page. Thanks.

Data Analyst
Explores traffic sources, conversion funnels, and retention, using clear metric definitions and concrete examples to explain what the numbers mean.