Customer and Partner Portal Development Cost: What It Actually Runs in 2026
A custom customer or partner portal costs $28,000 to $250,000, with most mid-market builds landing between $60,000 and $130,000. Timeline runs 8 to 12 weeks at the low end, 14 to 22 weeks in the middle band, and 6 to 9 months for anything with deep ERP (Enterprise Resource Planning) integration, compliance requirements, or a mobile app alongside the web build.
The honest bands, from 2,000+ Digital Heroes projects
Every portal quote is really a question about how many systems you are wiring together and how much of your business logic has to be re-expressed in software. Here is where builds actually land.
Tier 1: $28,000 to $55,000. Eight to twelve weeks.
This buys a web-only portal with login, role-based access for two or three user types, a dashboard, one or two core workflows (view invoices, submit a request, download documents), and a basic admin screen. Team is one designer part-time, one or two full-stack developers, and a part-time project manager. One integration, and only if that system has a modern REST API with decent documentation.
What falls out of scope at this price, and you should assume it does: no mobile app, no offline support, no data migration beyond a CSV import of user accounts, no single sign-on against your corporate identity provider, no SOC 2 or HIPAA work, no custom design system (you get a styled component library like shadcn or Material with your colors on it), no complex approval chains, and no ERP sync. If a vendor quotes $35,000 and any of those words appear in your requirements, the number is wrong and you will find out in month three.
Tier 2: $60,000 to $130,000. Fourteen to twenty-two weeks.
This is where most real portals live. You get four to eight distinct workflows, a proper account hierarchy (parent company with child locations and users), two or three integrations, a real admin back office your ops team can actually run without a developer, notifications, a custom design pass, and a migration of live historical data. Team is a designer, two or three developers, a QA engineer, and a project manager at roughly half time. This tier supports a few thousand users comfortably.
Tier 3: $130,000 to $250,000 and up. Six to nine months.
Deep ERP or legacy integration, four or more connected systems, a native mobile app in addition to web, compliance certification work, multi-tenancy with white labeling, real-time features, or a migration off a portal that already exists and has ten years of data in it. Team is five to eight people including an architect and a dedicated DevOps engineer. Above $250,000 you are usually building a product, not a portal, and the conversation should change accordingly.
What actually drives the number
Seven things move the price.
1. Integration count and integration age. $4,000 to $12,000 per system, or $15,000 to $30,000 if it is old. A modern REST API with OAuth and real docs (Stripe, HubSpot, Salesforce) costs $4,000 to $12,000 to wire up properly with error handling, retries, and a sync log. An on-premise ERP with a SOAP endpoint, a flat-file drop, or a database you are told to query directly costs $15,000 to $30,000 for the same functional result, because half the work is discovering what the data actually means. Two integrations is not double one integration. It is roughly 2.4x, because now you have conflict resolution between two sources of truth.
2. Compliance. Add 18 to 25 percent on the build, plus $15,000 to $40,000 outside it. HIPAA, SOC 2, or a serious enterprise security review adds audit logging on every data access, encryption at rest with key management, session policies, penetration testing, and documentation nobody enjoys writing. The 18 to 25 percent is engineering effort inside the project. The $15,000 to $40,000 is the external auditor and the pen test, which is not your development vendor's line item and is frequently missing from the budget people bring us.
3. Data migration. $6,000 to $35,000, driven by source quality, not record count. Migrating 400,000 clean rows out of a well-normalized database is cheaper than migrating 4,000 rows out of a spreadsheet where three columns mean different things depending on the year. Budget $6,000 to $10,000 if your data lives in one system with a schema. Budget $20,000 to $35,000 if the answer to "where does customer data live" takes more than one sentence.
4. Mobile alongside web. Add 55 to 70 percent, not 100 percent. The API, business logic, and database are already built, so the mobile app reuses the expensive half. What you pay for is a second set of screens, a second QA cycle, app store submission, push notification infrastructure, and the ongoing tax of two release trains. A responsive web portal that works well on a phone is free. A native app is not.
5. Design depth. $5,000 to $8,000 versus $22,000 to $40,000. Component library plus your brand colors plus sensible layouts is the low number and it is genuinely fine for an internal-facing dealer portal. A custom design system with a full component inventory, motion, empty states, and dark mode is the high number, and it is worth paying when the portal is customer-facing and competes with someone else's portal.
6. Real-time and offline. Add $12,000 to $30,000. Live status updates, presence, collaborative editing, or a field app that has to work in a basement with no signal all require different architecture from the start. Retrofitting real-time into a request-response portal after launch costs roughly double what building it in costs.
7. Scale and tenancy. Add $10,000 to $25,000. Multi-tenant isolation, granular permissions beyond three roles, and performance work for hundreds of concurrent users. Under 500 users this is close to zero. Above 5,000 with tiered permissions it is real money.
A worked example: partner portal for an industrial distributor
1,400 dealer accounts, tiered pricing, order history, quote requests, warranty claims, and a document library. Web only, no mobile app, no compliance certification. This is the single most common shape of request we quote.
- Discovery, scope definition, technical design: $7,500
- UX flows and wireframes, 16 screens: $6,000
- UI design system and high fidelity screens: $9,500
- Auth, SSO, roles, dealer account hierarchy: $9,000
- Dealer dashboard, order history, reorder flow: $13,000
- Quote request and internal approval workflow: $12,000
- Warranty claim submission with file upload: $8,500
- Document library with tiered permissions: $5,000
- Admin back office (dealers, pricing tiers, content): $12,000
- ERP integration, order and pricing sync: $16,000
- Salesforce integration, account and lead sync: $6,500
- Data migration, 1,400 accounts and 6 years of orders: $8,000
- Email and in-app notifications: $3,500
- QA and user acceptance testing: $10,000
- Infrastructure, CI/CD, environments, monitoring: $5,500
- Project management across 18 weeks: $9,000
- Launch, training, documentation, 30 days hypercare: $4,500
Subtotal: $145,500. Contingency at 10 percent: $14,550. Approved budget: $160,050.
The client walked in with a $90,000 number in their head, which is a reasonable Tier 2 guess. The ERP integration ($16,000), the admin back office ($12,000), and the approval workflow ($12,000) are what pushed it into Tier 3. Those three lines total $40,000 and none of them appeared in the original one-page brief. That is the gap, and it is the same gap almost every time.
The ongoing costs nobody puts in the quote
Hosting and infrastructure: $200 to $2,000 a month. A portal with a few thousand users on managed hosting with a managed database, staging environment, backups, and a CDN sits around $400 to $900 a month in practice. Document-heavy portals push toward the top of the range once object storage passes a few hundred gigabytes.
Third-party services: $300 to $2,500 a month. Identity providers like Auth0 and Okta price per monthly active user, and a portal in the 2,000 to 10,000 active user range typically lands between $300 and $1,500 a month. Add transactional email, SMS if you send it, error monitoring, and analytics. Each is small. Together they are a real line.
Maintenance: 15 to 20 percent of build cost per year. On the $145,500 example that is $21,800 to $29,100 annually. It buys dependency and security patching, framework upgrades, integration repair when your ERP vendor changes an endpoint, browser regressions, and the support hours when a dealer cannot log in on a Friday afternoon.
Year one changes: 25 to 40 percent of build cost. Once real users touch the portal, the business will ask for things. A new report. A field on a form. A different approval rule. A second document category. None of it is scope creep, it is the portal working. Budget $36,000 to $58,000 against the example above, or accept that the portal freezes on launch day and starts decaying.
How to not get burned on price
When a $145,000 build gets quoted at $70,000, one of three things is true. The vendor did not understand the scope and will discover it in week six, at which point you are already committed and the change orders begin. Or they understood it and priced to win, planning to make the margin back on variations. Or they are staffing it with people who have never built this before and you are funding their education. In our experience the pattern is consistent: the cheap quote lands at 1.4x to 1.8x its original number by the end, and it takes longer than the honest quote would have.
A change request should cost what it costs, transparently. Ask for the blended hourly rate in the contract and ask that every change request come with an hours estimate against that rate before work starts. A small change (a field, a filter, a new email template) should be 4 to 12 hours. A new workflow is 40 to 120. If a vendor cannot tell you the rate, or quotes changes as opaque lump sums, that is the mechanism they will use to make back a lowball bid.
Three contract terms that protect the number. First, fixed scope with a written scope document that both sides sign, listing what is in and what is explicitly out. Vague scope is how fixed price quietly becomes time and materials. Second, IP transfers to you on payment, per milestone, not at project end. If the relationship breaks in month four you still own months one through three. Third, source code lives in your repository from day one, with your CI/CD in your cloud account. If you cannot deploy the thing without calling the vendor, you do not own it, you rent it.
How to brief a vendor so the quotes come back comparable
Five vendors reading a one-page brief will each imagine a different project, and then you compare five numbers that describe five different things. Give every vendor the same document containing:
- User types and counts. Not "customers." Say: 1,400 dealer accounts, average 3 users each, plus 25 internal staff across two roles.
- Every workflow as a sentence starting with a verb. "A dealer submits a warranty claim with photos, an internal rep approves or rejects with a note, the dealer is emailed the outcome." Ten of these sentences is a scope document.
- The integration list with the actual system names and versions. "SAP Business One, on-premise, version X" produces a real number. "Our ERP" produces a guess.
- What data moves, from where, how much. Include a sample export. This single attachment changes migration quotes more than any other input.
- Your compliance and security requirements, stated up front. If your enterprise customers will send a security questionnaire, say so now.
- Your real budget range and your launch date. Withholding the budget does not get you a better price. It gets you a quote for a project you cannot afford, and three weeks lost.
Then ask every vendor to price the same line item breakdown rather than a single total. When the numbers come back you will be comparing arithmetic instead of confidence, and the vendor who misunderstood the project will be obvious from which line they underpriced.
The evidence behind this guide
Independent findings on why this investment pays off. Every link goes to the primary source.
- 73% of surveyed businesses now use a headless architecture (up nearly 40% since 2019), and 98% of those not yet using it are evaluating or planning to evaluate headless within 12 months, with 82% saying it makes delivering consistent content easier. Source: WP Engine (2024) →
- McKinsey emphasizes that most L&D functions still fail to tie training to business outcomes, recommending organizations track 2-3 business-relevant indicators (such as time-to-proficiency, redeployment into priority roles, or frontline productivity) rather than participation metrics to demonstrate training effectiveness. Source: McKinsey & Company (2025) →
- Per the Standish Group CHAOS 2020 report (reviewed at this URL), across tens of thousands of software projects roughly 31% end successfully, about 50% are 'challenged', and roughly 19% fail outright; small projects succeed far more often than large ones, and Agile approaches succeed at markedly higher rates than Waterfall. Source: The Standish Group (2020) →
- In an RCT, the no-show rate was 23.5% for patients receiving a text-message reminder versus 38.1% for the control group - a 14.6 percentage-point reduction (p = 0.04). Source: Clinical Pediatrics / PubMed Central (Lin et al.) (2016) →
Rohan advises mid-market and enterprise teams on ERP, CRM and custom software, and has led delivery on dozens of business-software builds.
Writes for Digital Heroes, shipping business software for 2,000+ brands across 55+ countries since 2017.