The CIAM Buyer’s Handbook is a free, seven-chapter ebook for the people who pick a customer identity platform: CTOs, product leads, security architects and the procurement partner who has to sign the contract. It covers scoring vendors, implementation and migration, and how to judge whether a vendor will still look the same in three years. You read it in the browser, with no signup and no email gate.
I wrote it because I spent more than a decade on the vendor side of this decision. I founded a CIAM platform, LoginRadius, in 2013 and scaled it to more than a billion users. The buyers who did well were rarely the ones with the longest feature checklist. They were the ones with a method.
CIAM (customer identity and access management) is the login, signup, consent and account layer your customers touch. Swapping it later means moving every user record and, ideally, every password hash. That is why the cost of a wrong pick shows up a year or two after the contract, not on day one.
The vendor list also moves under your feet. Three recent events make the point:
None of these were bad products. Each one still forced customers to re-plan. The handbook treats that risk as an evaluation criterion, next to security and price.
| Chapter | What you get from it |
|---|---|
| 1. CIAM 101 | Why customer identity is a product decision, a CIAM vs. workforce IAM matrix, and a scored build-vs-buy test. |
| 2. The Evaluation Framework | Five scoring dimensions, a weighted template, a must-have list, and 15 questions to put to every vendor. |
| 3. Solution Landscape | Market categories and ten vendor profiles with pricing checked in November 2026, from full platforms to enterprise SSO specialists, matched to company stage. |
| 4. Implementation Playbook | A five-phase, 14-week rollout plan, three migration strategies, and six common implementation mistakes. |
| 5. Decentralized Identity | The EU Digital Identity Wallet, verifiable credentials and what to ask vendors about them now. |
| 6. Open Source CIAM | Keycloak, Ory, ZITADEL, Authentik and FusionAuth compared, plus the real cost of self-hosting. |
| 7. CIAM Market Analysis | Who bought whom, how private equity ownership changes a vendor, and the contract clauses that protect you. |
Workforce IAM (Okta Workforce, Microsoft Entra ID) is built for employees: IT-provisioned accounts, internal apps, per-user pricing. CIAM is built for self-service signup, social and passwordless login, consent records and spiky traffic. Chapter 1 puts the two side by side across ten dimensions. Its short version is a decision matrix:
Many B2B SaaS companies need both. The mistake the chapter warns against is forcing one system to do both jobs.
Chapter 2 scores every vendor on security, scalability, developer experience, compliance and pricing. The weights change with the business, and the book shows two example weightings:
| Dimension | B2B SaaS, enterprise customers | High-volume consumer app |
|---|---|---|
| Security | 30% | 20% |
| Scalability | 15% | 30% |
| Developer experience | 20% | 20% |
| Compliance | 25% | 10% |
| Pricing | 10% | 20% |
In the chapter’s worked example, the same two vendors swap places depending on which weighting you use. Fix the weights first, and a polished demo cannot quietly reorder your priorities.
The chapter lists 15 of them, grouped by pricing, lock-in, security and reality. Five I would never skip:
The hash-export question matters most. If hashes cannot leave, every user resets a password when you migrate.
Chapter 4 describes three ways to move existing users onto a new platform:
| Strategy | How it works | Trade-off |
|---|---|---|
| Bulk import with hashes | Export password hashes and import them, so users keep their passwords. | Both systems must support the hash format. |
| Lazy (just-in-time) migration | Check each login against the old system, then create the account in the new one. | You run two systems until the cutoff. |
| Forced reset | Move everything except passwords and email every user a reset link. | In the book’s experience, a meaningful share of users never finish the reset. |
The practical lesson: a vendor’s import and export support is part of the price. Score it in the evaluation, not after.
Chapter 7 tracks the consolidation wave, from SAP buying Gigya to Thoma Bravo combining Ping Identity and ForgeRock. The buyer’s response it recommends has two halves. In the contract, negotiate:
In the architecture, put a thin identity abstraction layer between your services and the vendor SDK. Use OAuth 2.0, OpenID Connect and SCIM (the standard for syncing user directories) wherever you can. Then run a test export every quarter so the escape hatch is proven before you need it.
The handbook gives you the method. CIAM Compass is where I keep the data that goes into it, updated as vendors change. Use the vendor directory to build a longlist, then head-to-head pages such as Keycloak vs ZITADEL to settle close calls.
Two of its tools map straight onto the book. The TCO calculator runs the three- and five-year cost model that chapter 2 asks for. The RFP builder turns your must-have list into a document you can send to vendors. For market events like the Twilio and Stytch deal, the changelog records what changed and what it means for existing customers.
If your team is still debating which vendor fits your stage, my earlier post on the CIAM vendor selection trap covers stage fit. The handbook picks up where that post stops: scoring, contracting and cutting over.
Yes. Every chapter is readable in the browser at guptadeepak.com/ebooks/ciam-buyers-handbook/ with no signup, no email gate and no trial. You can link a colleague straight to the chapter they need.
The handbook’s process runs five weeks: define requirements and weights in week 1, shortlist in week 2, run proof-of-concept builds in weeks 3 and 4, and score in week 5. Compressing it is how teams end up migrating again within two years.
Usually not, unless identity is your product. Chapter 1 gives a scored test: rate five reasons to build and five reasons to buy from 1 to 5. Building only makes sense when the build score is 20 or more and the buy score is under 15.
Ask whether you can export all user data including password hashes, whether tokens are standard JWTs, how long a typical customer migration off the platform takes, and what happens to your data if you stop paying. Get the answers in writing.