What a Client Portal Actually Costs at 5, 20 and 50 Clients

The short answer: entry pricing is real but almost nobody stays on it. White-label and custom domain live on the second or third tier at most vendors, not the entry tier the marketing page uses as a hook. At 10 internal users the same requirement costs $19 a month on SuiteDash and $462 a month on Copilot Professional. And the platforms that price by client count rather than seat count get expensive in a way that scales exactly with your success.

This post models the real bill rather than the headline. Prices verified across vendor pricing pages and independent comparisons through mid-2026.

The two pricing models, and why they matter more than features

Every client portal falls into one of two shapes, and the shape determines your bill far more than the feature list does.

Flat-fee platforms charge a fixed monthly rate regardless of how many clients you add. SuiteDash, HoneyBook Starter, Agiled and Clinked work this way. The curve flattens until you hit a member or storage cap.

Per-seat or per-client platforms scale linearly. Bonsai, Notion, Copilot past its first tier, and project-management tools with guest access all charge more as you grow.

The trap is not that per-seat pricing is more expensive in principle. It is that adding a client is the thing your business is trying to do, and one of these models charges you for succeeding.

Entry pricing, verified

Butterbase appears here as a build / fork option rather than a finished portal product: you start from a working client-portal template and modify it, so the cost shape is flat and client count does not move the bill.

Those are the numbers on the pricing pages. Now the parts that are not.

The tier traps

HoneyBook puts its own branding on every client-facing document. Removing it means moving up a tier. There is no custom domain option at all.

Copilot caps Starter at 50 clients for $39 a month. Professional runs $89 a month for up to 500. Custom domain support requires the Advanced plan at $149 a month. And internal users past the first three cost $39 each.

Plutio gates white-label behind its Max plan.

Clinked's entry tier is $95 a month, and the compliance-ready Premium tier with SOC 2 and enforced two-factor runs $383.20 a month billed annually. Most "best client portal for small business" listicles quietly leave that number off.

Moxo starts at $200 a month per its 2026 pricing page, which already exceeds most alternatives at ten seats before you have configured anything.

One analyst who modelled a five-person team needing a white-labelled portal on a custom domain with payment processing and branded email put it plainly: the tier traps turn a $19 a month portal into a $159 a month bill the day you need single sign-on or a fully white-labelled mobile app.

The 10-user comparison

Here is the number that reframes the category. For the same ten internal users:

A 24x spread for the same headcount. The pure-portal tools, Clinked and Moxo, start at a price that already exceeds SuiteDash or Agiled at ten seats before you add anything.

What happens as clients grow

Project-management tools with guest access are the sharpest version of the problem. On annual list pricing for a five-person internal team with interactive client access, Monday Standard runs $864 a year at five clients and crosses $1,700 a year at thirty. ClickUp can stay cheaper by moving to Business, around $720 at five seats, once you outgrow Unlimited's guest allocation.

Many tools count clients as paid seats once you add them. Flat-fee platforms like SuiteDash and Rock do not, which is the cleanest cost story in the category.

The practical test: ask any vendor what your bill looks like at three times your current client count. If they cannot answer in one number, you are on a per-client model whether or not the pricing page says so.

The bill at 5, 20 and 50 clients

Putting it together for a ten-person firm that needs white-label and a custom domain - the realistic minimum for anything client-facing:

The flat-fee row is the only one where growth is free. The per-seat and per-client rows rise in a straight line with exactly the metric you are trying to grow. And the build row shows why "just build it" is wrong advice below roughly twenty clients: the amortised cost dwarfs every subscription on the table.

These figures use published list pricing and will drift - but the shape will not. Verify the current number with any vendor before signing, and always with the same question: what is my bill at three times my client count, on the tier that has white-label?

The costs that are not on any pricing page

Setup time. A portal takes one to three weeks to configure properly - branding, permissions, templates, migrating active clients. That is internal time no invoice shows.

Migration. Moving clients off a platform you have outgrown means re-inviting every client, re-linking every document, and a transition period where two systems run in parallel. Budget a week of someone's time per platform switch.

The second subscription. The tool you buy rarely replaces the ones it overlaps. Most firms end up paying for the portal and the project tool and the file storage the portal was supposed to absorb, at least for a year.

Renewal drift. Annual pricing is the number on the page; monthly billing typically runs 20 percent higher, and second-year renewals are where quiet increases land. Put the renewal date in your calendar the day you sign.

The build option, and its real numbers

The other path is building your own, and the numbers quoted for it are consistently large. Custom client portal development is quoted at $20,000-$60,000 upfront, with $10,000-$25,000 a year in maintenance. Developer rates of $45-$200 an hour, plus the ongoing cost of a system nobody on the team owns after the contractor leaves.

The argument for building is real: no per-seat fees, exactly your workflow, and you own it permanently. For businesses with 20 or more active clients, custom becomes more economical over a long enough horizon. The argument against is equally real: three to six months to first launch, and a maintenance obligation that does not end.

The fourth option

Buy, build, or subscribe are the three options every comparison lists. There is a fourth, which is to fork something that already works and run it yourself.

A forked application starts as working software rather than a blank repository. Auth, permissions, tenant separation and audit already exist, because they are properties of the platform underneath rather than things each project rebuilds. You modify what is specific to your business and skip what is not.

The cost shape is different from all three: no per-seat pricing, because clients are not seats; no per-client pricing, because the boundary is a database policy rather than a billing tier; no $20,000 build, because you are not starting from nothing; and self-hostable, which matters for the compliance cases.

Where this does not apply: if you need a portal live this afternoon and have five clients, SuiteDash at $19 flat is a genuinely good answer and you should take it. The fork option earns its place when you are either past the point where per-client pricing hurts, or you need something the off-the-shelf products do not do.

How to actually decide

Model three client counts, not one. Your current number, three times it, and ten times it. If the curve bends, you are on a model that charges you for growth.

Check where white-label and custom domain sit. They are almost never on the entry tier, and they are almost always the first thing a client-facing portal needs.

Count internal users separately. Copilot charges $39 per internal user past three. That is a different bill from the one on the pricing page.

Ask what happens at renewal. Annual pricing is often quoted; monthly is typically 20 percent higher.

And ask the question that decides it: are you paying per client? If so, every new client makes the software more expensive at exactly the moment it should be paying for itself.

Butterbase takes the fork approach. Clone a production-ready client portal, run it as your own, self-hosted if that is the requirement. Tenant isolation is enforced in Postgres rather than in application code, so external client access is a database guarantee rather than a pricing tier. No per-seat pricing, and no charge for the clients you add. Related reading: The Compliance Surprise That Kills Client Portal Projects.

Frequently asked questions