Client portals are frequently commissioned and rarely used. The ones that work share a single characteristic: they answer a question faster than emailing you would.
Everything else about the design is secondary to that.
Why most portals fail
They are built to save the business time. A portal that exists so customers stop ringing to ask about progress is solving your problem, and customers are under no obligation to help.
Consider it from their side. They have a question. Option one is to send you a two-line email from the phone in their hand. Option two is to find the portal address, remember a password, reset the password because they have not logged in since March, navigate to the right section, and read the answer.
Option one wins unless option two is genuinely faster. That is the whole design problem.
Where portals genuinely earn their place
Repeat customers with frequent questions
Somebody who deals with you weekly will learn a portal, because the time saved compounds. Somebody who buys once a year never will.
Information that changes and matters
Live job status, delivery tracking, ticket progress, current stock levels for a trade account. Information with a state that customers actively wonder about.
Static documents alone rarely justify a login. If it never changes, email it.
Documents customers need repeatedly
Certificates, test reports, service records, past invoices. The value here is retrieval — customers lose these constantly and asking you for a copy is embarrassing for them.
This is often the single most appreciated feature and it is also the simplest to build.
Anything with a compliance dimension
Where a record of who accessed what and when is needed, a portal is far better than an email trail.
Trade and account customers
Business buyers with credit accounts, agreed pricing and repeat orders are the strongest case there is. See selling to trade customers online.
What to build first
Look at your inbox. Find the three questions customers ask most often. Build the answer to those three, and nothing else.
For most service businesses that is:
- Where is my job or order up to?
- Can I have a copy of that document?
- What do I owe and what have I paid?
Three screens. Resist everything else until those three are being used daily.
Portals fail from breadth, not depth. A portal doing three things well is used. A portal doing eleven things adequately is not.
The parts people underestimate
The integration is the project
The portal itself is straightforward. Getting live job status out of your operational system, invoices out of your accounting package and documents out of wherever they live is where the effort goes.
If the data is not already in a system that can be read programmatically, the portal cannot show it. That constraint sets the scope, and it is covered in how integrations work.
Getting people in
Password resets are the largest source of portal support work. Consider magic links sent by email instead of passwords — the customer clicks a link and is logged in. Fewer forgotten credentials, less support, better adoption.
Permissions
Customer A must never see customer B's data. This sounds obvious and it is the single most common serious defect in portals, because permissions must be enforced on every request rather than by hiding links.
It must work on a phone
Portals are checked on phones, in vans, on sites, between meetings. A desktop-only portal will not be used.
How to get it adopted
Adoption is a process, not a launch.
Answer questions with links. When a customer emails asking about status, reply with the answer and a direct link to where it lives. Repeat for a few months.
Send links, not attachments. Invoices and certificates delivered as portal links teach the habit gently.
Make it faster, visibly. If the portal shows something they cannot get any other way — live tracking, real-time stock — it sells itself.
Do not force it. Removing the email option to drive adoption produces resentment, not usage.
When not to build one
If your customers buy once, if the information rarely changes, if you have fewer than about twenty active accounts, or if the underlying data is not in a readable system — a portal will not repay its cost.
In those situations a better-automated email is usually the right answer, and considerably cheaper. Our automation guide covers where to start.
Our web applications team builds portals for UK businesses and will tell you when a well-designed email would serve you better. Start a conversation.
