Client Portal Software, Build or Buy: Where Ready-Made Portals Break

Somewhere around the thirtieth active client, the email threads stop working. A client asks where their documents are. Someone forwards the wrong version. An invoice goes unpaid because it was buried under a status update. You search for client portal software, find forty products that all look fine in the demo, and hit the real question: buy one of these, or build something that fits how you actually work?

This post is about portals specifically: document sharing, job or case status, invoices, intake forms. The general framework for any software decision lives here; this post applies it to one category where the trade-offs are unusually sharp.

What off-the-shelf portals do well

Be fair to the products first. For a business with a simple, repeatable client relationship, a ready-made portal is close to a solved problem.

They launch in a week, not a quarter. Sign up, upload a logo, connect payments, invite clients. Going from email chaos to "everything in one place" is a big step, and these tools get you there fast.

The basics are good. Secure file sharing, e-signatures, invoicing with card and bank payment, intake forms, a message thread per client. Mature portal products have polished all of these because every customer needs them.

The price is low at small scale. SuiteDash charges a flat $19 to $99 a month with unlimited staff and clients. HoneyBook runs $29 to $109 a month on annual billing, plus card fees of 2.7 percent and 10 cents per payment. Assembly (formerly Copilot) has a free tier for one user and five contacts, and a Starter plan at $29 a month billed annually. For a solo consultant, these numbers are a rounding error.

Someone else handles security and uptime. Password resets, encryption, backups. None of it is your problem.

If your clients all go through the same four steps, your staff count is small, and a logo is enough branding, stop reading and buy one. The rest is for businesses that did that and hit a wall.

Where they break: branding and the client's experience

The first wall is cosmetic, and it matters more than it sounds. Clients log in to a URL that isn't yours, see a layout you didn't choose, and get emails with someone else's badge at the bottom. On cheaper plans this is normal. Assembly reserves custom domains and badge removal for its Professional tier and above, and HoneyBook lists removing its "Powered by" mark as an Essentials-plan feature. SuiteDash advertises heavy white-labeling even on its entry plan, one reason agencies pick it.

The real cost isn't the upgrade fee. It's that the portal is often the only software your client touches with your name on it. For a law firm or a home-care agency, that screen is the experience between meetings. When it looks like a generic SaaS tool, the client's sense of what they're paying for drifts down with it.

Where they break: workflow fit

The second wall is structural. Every off-the-shelf portal encodes a workflow: a client, a project, files, invoices, tasks. If your business runs on that shape, great. Many don't.

A home-care agency needs a portal where the client is a family, the "project" is a schedule of visits by different caregivers, and the documents are care plans that change weekly. A contractor needs change orders tied to line items in a quote, with photos attached. An immigration firm needs intake that branches twenty ways by visa type. A clinic needs intake that writes into the patient record, not into a PDF nobody re-keys.

You can bend a generic portal toward each of these with custom fields and naming conventions. You'll spend hours doing it and keep spending them, because every new hire learns the workaround and every client sees a confusing version of your process. The telltale sign is a spreadsheet kept beside the portal "because the portal doesn't track that." You're paying for software and still doing the work by hand.

Where they break: data ownership and integrations

Third wall. Your client records, documents, messages and payment history live in someone else's database. Most products let you export, but exports are flat files, and the relationships between things rarely survive the trip. Switch tools and you're rebuilding history by hand or abandoning it.

Integrations are the related problem. A portal that doesn't talk to your scheduling, accounting or CRM means status gets updated twice, or not at all. Vendor connectors and Zapier handle one-way pushes fine. For anything two-way, or anything that must be right every time (invoice totals, appointment changes), the glue becomes its own maintenance burden. Several of the automation examples here began as exactly this gap between a portal and the systems around it.

Where they break: per-seat pricing at scale

The fourth wall is arithmetic, and it's the one that makes the build conversation serious.

Entry prices are low because they assume few staff and few clients. Assembly's Starter plan includes one internal user and 50 active contacts; Professional and Advanced include three and five users, with more seats costing extra. HoneyBook's Essentials plan caps you at two team members and sends you to Premium for unlimited. Flat-priced tools like SuiteDash avoid the per-seat trap but cap storage (100GB on the entry plan, extra billed by the block), and on every ladder the features a growing firm needs (white-label, custom domain, API, automations) sit on the higher rungs.

Run the numbers three years out, with the staff and client counts you expect, on the tier that has what you need. For a 15-person firm with 400 active clients, $200 to $600 a month is common once you add seats, storage and add-ons. That's $7,000 to $22,000 over three years, for software that still doesn't quite fit. Not an outrage, just a real number to compare against building.

When a custom portal pays for itself in hours

The comparison that matters isn't software cost. It's staff time.

Add up the hours your team spends each week on portal workarounds: re-keying intake data, answering "where's my document" emails, updating status in two places, maintaining the shadow spreadsheet. Then multiply by a loaded hourly cost. The US median annual wage for office and administrative support roles was $47,450 in May 2025, roughly $23 an hour before benefits and overhead, so $30 to $35 loaded is reasonable, and higher if the hours belong to a professional rather than an admin.

A firm losing 15 staff hours a week to portal friction loses around 750 hours a year, which at $32 an hour is $24,000, or $72,000 over three years, before counting clients who got a worse experience. If a custom portal costs less than that to build and run, and removes most of those hours, it pays for itself. At three hours a week, it doesn't; stay on the off-the-shelf product and accept the friction.

The hours are also why a portal built and maintained over years by people who understand your operations is a different asset from a one-off project. The hours it returns are the return.

What a minimal custom portal actually includes

"Custom" scares people because they imagine a two-year enterprise project. A useful portal for a growing service firm is a bounded, well-defined piece of software. The minimum is roughly this:

• Login with roles. Clients see their own things, staff see everything.
• A client record that matches your business. Matter, job, patient, household, with the fields you actually track.
• Document upload and sharing, with versioning and a clear "what we need from you" list.
• Status your team updates once and the client sees instantly, ideally fed by your existing systems.
• Invoices and payment through Stripe or similar, linked to the specific job.
• Intake forms that write into the client record, with the branching your process needs.
• Notifications by email (and SMS if clients prefer it) when something changes or is due.
• Integrations with the two or three systems you already run, so nothing is entered twice.

Leave out at first: built-in messaging (email works), a mobile app (a responsive web app does the job), and every "it would be nice if." Add those once the portal has proven it saves hours.

On a modern web stack with a managed database and auth provider, this is a bounded build with predictable hosting costs, and it's yours: your data, your domain, your workflow, and no seat count to argue about when you hire.

A short decision checklist

Buy an off-the-shelf portal when: your client workflow is simple and uniform, your staff count is low and will stay that way, a logo is enough branding, and the portal doesn't need to talk to other systems.

Consider building when at least two of these are true: staff keep workarounds or shadow spreadsheets around the portal, your process doesn't fit the vendor's shape, clients regularly see the vendor's brand instead of yours, your three-year subscription on the right tier runs into five figures, or the portal needs two-way integration with core systems.

Either way, write down the hours before you decide. The hours are the honest number.

Next step

If you're weighing a portal subscription against building your own, we're glad to look at your current process and your numbers and tell you which way we'd go, including when the answer is "stay on the $49 plan and fix the two workflows around it." Book a free call and you'll get a written recommendation within a few days, whether or not you work with us.