Type "client portal builder for marketing agencies" into Google. Every single result on page one is a SaaS subscription product — Moxo, AgencyAnalytics, Clientjoy, Noloco, Clinked, Knack, Assembly, ClientPortal.io, AgencyHandy, DaxRM. They're priced per seat, billed forever, and locked to whatever roadmap their product team feels like shipping next quarter.
Custom-built portals don't appear in those results. Not because they're worse — for the right agency they're dramatically better — but because nobody who builds custom invests in ranking for the comparison query. So the buying signal is broken. You search for "the right tool" and the only tools you see are the ones that treat your agency like every other agency on Earth.
This article fixes that. Below is the actual framework we use at AppBox when an agency owner asks us whether they should pay $79 a seat for SuiteDash or commission a custom portal that lives on their own domain forever. The honest answer depends on two questions, and most agencies get them wrong.
The two questions that decide everything
Strip away the feature lists and the demo videos. Strip away the screenshot grids and the "trusted by 5,000 agencies" badges. The choice between SaaS and custom comes down to two things:
- Question 1. Does your agency's workflow fit the shape of a generic agency workflow — or have you built something that requires its own logic?
- Question 2. Are you willing to spend the next five years absorbing someone else's product roadmap, pricing changes, deprecations and "we're sunsetting that feature" emails?
If you answer yes to question one and yes to question two — SaaS is the right call. Stop reading. Buy SuiteDash, Moxo, or Clinked, and get on with running the agency.
If you answer no to either, you're the type of agency we built AppBox for. Read on.
Where SaaS portal builders genuinely win
We don't think SaaS is bad. We think SaaS is mismatched for a specific kind of agency, and a great fit for another. Here's where the SaaS portal pattern is the right answer:
1. You're under five people and your services are templated
If you sell social media management at a fixed monthly retainer, an SEO audit at a fixed price, or three-package design retainers — your workflow is generic by design. SaaS portals were built around exactly this shape. You'll get an onboarding form, a deliverable folder, an approvals UI, an invoicing module, and a chat thread. That's enough. Don't over-engineer.
2. You don't need to integrate with anything weird
Most SaaS portals integrate cleanly with Stripe, Xero, Google Calendar, Slack, and a dozen marketing-stack tools. If your stack is "the obvious stack," that's all you need. The integrations are pre-built, they're maintained by someone else's team, and they survive vendor API changes without you noticing.
3. You want to spend zero hours thinking about software
A SaaS portal arrives with auth, billing, file storage, audit logs, GDPR responses, password resets, backup policies, status pages, support agents, and a release schedule already wired up. You don't have to maintain any of that. For a small agency that's an enormous gift.
Where SaaS portal builders fall apart
The further your agency moves from "generic agency" toward "we've built a real specialism," the worse SaaS fits. Below are the moments we've watched it break, repeatedly, with founders we work with across Australia, the US, the UK and Canada.
1. Your workflow doesn't match the SaaS object model
Every SaaS portal is built on a fixed object model — usually Client → Project → Task → File → Comment. That's beautiful when your work fits the model and a knife fight when it doesn't. Promo merch agencies, for instance, don't sell projects — they sell quotes, which become orders, which spawn production runs, which trigger artwork approvals, which feed shipping batches. Try cramming that into a generic project tool and you'll spend a year retraining staff to lie to the software.
2. You need real, two-way integration with your real stack
SaaS integrations look great in the marketing site and feel thin in real life. "Integrates with Xero" usually means a one-way export of invoices. "Integrates with HubSpot" usually means a contact sync that breaks every 90 days. When the integration you actually need is "my client's approval inside the portal must trigger a production run in our supplier API and a reorder threshold check in our Notion ops board", you've left the surface area SaaS covers.
3. The portal is the brand experience, not just admin
For a lot of agencies, the portal is the most-used surface their client touches. It's where the relationship actually lives. SaaS portals all look generic SaaS — sidebar, top bar, white card, blue button. If your agency's positioning is premium, design-led, or distinctly opinionated, putting clients into a generic UI undercuts everything you sold them in the pitch deck. (For the parallel argument on agency websites themselves — and why every promo merch site looks the same — see our promo websites guide.)
4. The per-seat math gets ugly fast
$79 per seat per month feels reasonable when you're three people. At 18 staff plus 40 client logins, you're paying $4,500 per month — $54,000 per year — forever. We've worked with agencies who spent more on portal subscriptions in three years than a custom build would have cost outright.
Where custom client portals win
Custom portals stop making sense as soon as your agency is small and generic. They start making extraordinary sense the moment you're not. Here's where we see custom pay back in twelve to eighteen months, repeatedly:
1. You're in a non-generic vertical
Promo merch (see our supplier API integration guide for that pattern in detail), print, branded apparel, sports rights, wholesale FMCG, B2B procurement, regulated industries, health, legal, government — none of these fit the generic agency object model. They have artwork approval chains, supplier APIs, MOQs, certificate trails, compliance fields, regional pricing, batch numbers. SaaS can't model that without you bending the workflow to fit. Custom can model it exactly.
2. You're already paying for integrations that don't really work
When the agency owner can describe — by name — three integrations that "kind of work but the team uses spreadsheets anyway because the Zapier flow falls over once a week," you're past the point custom would have paid back. Wiring directly into the source systems with a maintained custom integration removes a whole class of operational pain.
3. The portal is a strategic asset, not a cost
A custom portal is yours. It lives at portal.youragency.com. It compounds with your brand. It carries your design language, your tone, your IP. When you sell the agency, the portal goes with the business — not with the SaaS provider. We've seen agency exits where a defensible client portal added materially to the multiple.
4. You want fixed, declining cost — not a per-seat tax that grows with your team
The economics flip. A custom build is a one-time setup plus a small monthly hosting and maintenance fee. Add five staff next year — same hosting bill. Add fifty client logins — same hosting bill. The marginal cost of growth approaches zero. With SaaS, the marginal cost of growth is the cost.
Where custom client portals don't make sense
We're not romantic about custom. We tell agencies to walk away from a custom build all the time. The honest list:
- You don't have a defined workflow yet. If your operations are still being figured out, you don't want them frozen in software. Use a SaaS portal as scaffolding, learn the real workflow, then commission custom in 12 months.
- You're under five staff with no integration needs. The economics don't pay back fast enough. SaaS is right.
- Your budget is genuinely under $3,000 USD. A real custom portal isn't a $1,000 freelance project — it's a platform. If the budget isn't there, SaaS is far better than half-built custom.
- You're not sure the agency itself is the long-term plan. Custom infrastructure is for businesses that intend to compound. If you're testing whether you even want to keep running the agency, don't build the portal.
- You want to experiment with marketing automation, not run client work. The patterns are different — automation tools are SaaS for a reason.
A 5-question decision framework
Run your agency through these five questions. Score yes / no on each.
- 1. Workflow shape. Does your service involve any object that isn't a generic project or task — quotes, orders, artwork rounds, certifications, MOQs, batch numbers, regional inventory, supplier APIs?
- 2. Integration depth. Do you need at least one two-way integration with a system that isn't a top-20 SaaS — your supplier's API, your custom Notion stack, an internal ops database, a regulator's portal?
- 3. Brand experience. Is the portal a meaningful part of how your clients experience your brand, not just an admin surface?
- 4. Scale economics. Will you have more than ~15 staff or ~30 active client logins inside the next 24 months?
- 5. Time horizon. Do you intend to still be running this business in five years?
Three or more yes answers — custom is the right answer. Two or fewer — SaaS will serve you well. One yes and one strong "we're not sure" — book a discovery call and let someone honest pressure-test the numbers before you spend.
The real 3-year cost comparison
Let's get specific. Take a 12-person agency with 25 active client logins. Compare three options across three years.