Claude Design Client Portal Mockups Without Vendor

Claude Design for client portal mockups is the build-vs-buy question most solo and small-firm practitioners face but rarely think through carefully. Anthropic launched Claude Design on April 17, 2026, and the tool changes the math on portal mockups specifically. The traditional path: rent a client portal from MyCase, Clio, Smokeball, or PracticePanther, accept whatever client journey the platform supports. The new path: prototype your own portal in Claude Design, deploy via Claude Code if the prototype validates, hand off to a developer if it doesn't. For most firms with standard intake-and-document-collection workflows, the rented portal is still the right call. For firms with specific journeys that don't fit any vendor's defaults, the build path is now realistic in a way it wasn't 90 days ago. This walks through when each path makes sense, the actual mockup process, and the integration questions that determine whether build or buy wins. Pricing pulled from the Anthropic pricing page and the published pricing of comparable practice-management vendors.

When to build versus when to rent

Most firms should rent. The category exists because most firms have standard client-journey patterns: intake form, document upload, secure messaging, document signing, payment, matter-status updates. MyCase ($49-$129/seat/month), Clio Manage ($49-$139/seat/month for basic-to-essentials, with portal included or as add-on), Smokeball, and PracticePanther all handle these patterns turnkey. For firms whose client experience matches what the platform supports, the rental is the right call.

Rent when:

- Your client journey is standard (intake → document collection → e-signature → ongoing matter updates → final delivery). - Your firm has 1-15 attorneys and you don't have IT or design resources to build and maintain custom infrastructure. - You're already paying for a practice-management platform that includes a client portal. - You need built-in compliance certifications (SOC 2, HIPAA-ready) without doing your own audit. - You bill matter-related work on the same platform, having intake, time tracking, billing, and payments in one tool reduces operational complexity.

Build when:

- Your client journey doesn't fit any vendor's defaults. Specific examples: contingency-fee personal injury with multi-stage document collection over 6-18 months, immigration practice with USCIS-tied document workflows, estate planning with multi-document signing sequences, contract-drafting practices with client-collaboration on document drafts, multi-language practices needing branded localization. - You want a portal that feels like an extension of your firm's brand, not a vendor's branded interface with your logo in the corner. - You have IT or developer support (or can layer in Claude Code) to handle deployment and maintenance. - You're handling specific data-residency, encryption, or storage requirements that vendors don't support out of the box. - You've outgrown vendor portal capabilities and need workflows the platform can't add for you.

The Claude Design for legal operations 2026 anchor frames this as part of the broader build-vs-buy pattern. The solo practitioner intake build covers the simpler upstream case where building wins more clearly.

What a client portal actually needs to do

Before mocking anything in Claude Design, get the spec right. A modern client portal needs:

- Authentication. Client login with secure credentials, ideally with two-factor authentication. SSO via the client's existing identity (Google, Microsoft) is a nice-to-have for B2B clients. - Matter status view. What's the current state of the matter, what's pending from the client, what's been completed. Avoid attorney-jargon labels, use language clients understand. - Document upload and viewing. Clients upload documents, receive documents, see version history. Bulk upload, drag-and-drop, file-type validation. Documents stored with proper access controls and encryption. - Secure messaging. Threaded conversations between client and attorney team. Read receipts, attachment support, search history. - Payment integration. Invoice viewing, payment processing (Stripe, LawPay, or equivalent), payment history. Client trust accounting compliance for jurisdictions that require it. - E-signature integration. DocuSign, HelloSign, or built-in signing. Audit trail. - Notification preferences. Email, SMS, in-portal, clients pick what they want. - Mobile experience. Most clients access portals from phones. Mobile-first design is non-negotiable.

For a custom build, this spec is meaningfully larger than a typical internal-tool spec. Claude Design handles the front-end well; Claude Code handles the back-end. But the integration to e-signature, payments, and identity providers represents real engineering work, typically 2-6 weeks for a deployed portal versus a few days for a simpler internal tool. The how to use Claude Design in Claude Code legal workflow covers the broader workflow.

The mockup-and-validate process

Whether you ultimately build or rent, mocking the portal in Claude Design first is the highest-leverage move. The mockup gives you a concrete artifact to evaluate against vendor offerings, does the vendor's portal support what you actually need, or are you compromising on workflow because you couldn't see the alternative?

Step 1: Prototype the dashboard. First Claude Design prompt: *"Build a client portal dashboard for a [practice area] firm. Show: matter overview with status, pending actions for the client, recent messages, recent documents, payment status. Match the firm's brand tokens."* Iterate until the dashboard reflects your client journey.

Step 2: Prototype the document workflow. Second prompt: *"Build the document section. List of documents grouped by matter phase, upload area with drag-and-drop, view/download/sign actions, version history per document."* Refine until it matches your firm's typical document collection sequence.

Step 3: Prototype the messaging and notification flows. Third prompt: *"Build the messaging section. Threaded conversations, attorney team avatars, attachments, search."* Add notification preferences as a fourth screen.

Step 4: Validate with stakeholders. Share the prototype URL with a few existing clients (not all of them — pick representative ones who'll give honest feedback). Ask: would this be better than what you currently use? What's missing? What's confusing? Most validation conversations surface 3-5 specific changes that materially improve the design.

Step 5: Compare against vendor offerings. With the mockup as your reference, evaluate vendor portals (Clio, MyCase, Smokeball, etc.) honestly. Does any vendor's portal already support your workflow as designed? If yes, the rental wins on cost and operational simplicity. If no, build the portal in Claude Code with the validated mockup as the starting point.

Most firms doing this exercise discover the vendor portal supports 70-90% of what they need. The remaining 10-30% is sometimes a deal-breaker, often not. The mockup forces the conversation explicitly instead of leaving it implicit. The Claude Design system integration for law firm brand covers the brand-tokens setup that makes the mockup look like your firm.

Cost comparison — build vs rent over 24 months

Representative budget for a 5-attorney firm with 100-300 active client matters annually:

Rent path (Clio Manage with included client portal). Clio Essentials at $99/user/month annual = $495/month for 5 seats = $11,880 over 24 months. Includes client portal, matter management, billing, time tracking, payments, document storage. Setup cost: minimal, typically a 1-2 week onboarding. Ongoing cost: subscription only.

Build path (custom portal via Claude Design + Claude Code). Claude Team Standard at $25/seat/month annual for 5 seats = $125/month, $3,000 over 24 months. Initial portal build: 2-6 weeks, typically $5,000-$25,000 in developer or contractor time if not handled in-house, or 30-100 hours of internal IT time. Hosting cost: $50-$300/month, $1,200-$7,200 over 24 months. Integration costs (Stripe, e-signature, identity provider): $0-$300/month, $0-$7,200 over 24 months. Maintenance: 4-8 hours/month of IT time.

Rent vs build over 24 months for a 5-attorney firm: Rent typically $12-$15K total. Build typically $9-$40K total, depending on developer support. Build wins on cost only when developer support is in-house or low-cost; rent wins on operational simplicity and lower upfront commitment.

The second-order math: rent has predictable monthly costs and known capability. Build has variable upfront costs but eliminates per-seat scaling — adding the 6th attorney to the portal costs nothing additional (versus another $99/month on Clio). For firms expecting to grow significantly, the build math improves over time. For firms that are stable in size, rent typically wins.

The broader pattern: don't treat this as ideological. Rent when rent works. Build when build solves a problem rent can't. The Claude Design pricing tier breakdown for legal covers the per-firm-size math in detail.

The Bottom Line: The verdict: for most solo and small firms, the rented client portal (Clio, MyCase, Smokeball) is still the right call — operational simplicity, predictable cost, built-in compliance. Build when your client journey genuinely doesn't fit any vendor's defaults. The Claude Design mockup process is valuable even if you ultimately rent: it forces explicit comparison between what you actually need and what vendors offer. Many firms find 70-90% fit; the remaining gap may or may not justify building. Don't build because building is exciting; build because the rent options genuinely don't fit.