The uncomfortable truth about portals
Most client portals fail. Not technically; they run fine. They fail because customers log in once, poke around, and never come back. The firm pays for the build or the subscription, the customers keep ringing the office anyway, and after six months everyone quietly agrees not to mention the portal again.
This is not an argument against portals. A portal that customers actually use is one of the highest-value things a small firm can run, and I will show the retention arithmetic below. It is an argument for building the boring version. Because the gap between used and ignored portals is not budget or design polish. It is that used portals answer the questions customers actually have, and ignored portals showcase features the firm wanted to show off.
What customers actually use
Watch real usage on any small-firm portal and the same four things dominate, in roughly this order.
Job status. The single most-viewed thing, by a distance. "Where is my job up to?" is the question behind most inbound calls a trades or project firm gets. When a customer can open a page and see "materials ordered, first fix booked for the 14th", they do not ring to ask. This one screen carries the entire case for having a portal. It has to be honest and current, which means it must update itself from your real job data. A status page someone has to remember to update becomes a status page that lies, and one wrong status costs more trust than no portal at all.
Documents. Quotes, invoices, contracts, and certificates: Gas Safe records, EICR certificates, warranties, building control sign-off. Customers do not browse documents for fun. They need them at specific moments: remortgaging, selling the house, an insurance claim, a tenant dispute. A portal that reliably has the certificate two years later, without an email archaeology dig on either side, earns quiet loyalty. It also kills the "can you resend the invoice" emails that eat office time.
Approvals. One button: approve the quote, approve the variation, sign off the stage. Every day a variation waits for approval is a day the job slips. Making approval a tap on a phone, with a record of who approved what and when, shortens jobs and settles disputes before they start. That timestamped record is worth more than most owners expect the first time a "I never agreed to that" conversation happens.
Payments, at a distance behind those three. A pay button next to the invoice gets used and shortens the wait to be paid. It is worth having, but it is a convenience, not a reason anyone logs in.
What they ignore
The features that sink portal projects are almost always the impressive ones.
Dashboards full of charts. Customers do not want analytics about their own bathroom refit. Firms add charts because charts demo well. Nobody looks at them twice.
In-portal messaging. Your customer already has email and WhatsApp. A third inbox they must remember to check is a feature they will not use, and worse, messages sent there and missed become "why didn't you reply?" Use the portal to show state, and let conversation stay where it already lives.
Community, news, and content sections. No customer of a six-person M&E firm wants that firm's news feed. Every ignored section teaches customers the portal is not for them.
Accounts and passwords per project. Every login hurdle halves usage. One login per customer, a magic link by email rather than another password, and everything works properly on a phone, because that is where customers will open it. If any feature needs a desktop to work, most customers will never see it.
The pattern: customers use a portal to get answers and complete actions. They ignore anything that asks for their attention rather than saving it.
The retention arithmetic
Portals get pitched as a premium look. The real return is quieter and shows up in two places.
First, interruptions. Count the inbound "any update?" calls and emails your office handles in a week, and the minutes each one takes to answer, because answering usually means going and finding out. For a firm running 20 or 30 live jobs this is commonly several hours a week of pure interruption. A trusted status page removes most of it, not because customers care less, but because they can check without catching you.
Second, repeat business. Customers who can see progress feel in control, and customers who feel in control stay. Most disputes in trades and project work are not really about workmanship; they are about surprise. The variation nobody remembers agreeing, the delay nobody mentioned. A portal that shows state and records approvals removes the surprise, and with it most of the friction that stops a customer coming back or recommending you. When the certificate is still sitting in their portal two years later, you are also the firm they can actually find when the next job comes up.
If you build one, build it in this order
- Job status first. Wired to your real job data, not manually updated. If you do nothing else, this pays for the project.
- Documents second. Everything the job produces, filed to the job automatically, kept available after the job closes.
- Approvals third. Quote, variation and stage sign-off with a timestamped record.
- Payments fourth. A pay link on the invoice.
- Then stop. Resist everything else until customers ask for it. They almost never do.
One warning that decides success more than any feature: a portal is a window onto your operational data, not a separate thing to keep updated. If your jobs live in spreadsheets that are themselves out of date, the portal will faithfully display wrong information to your customers. Firms usually discover this the other way round: they come asking for a portal and find the real project is the job tracking underneath it. If that sounds familiar, start with the spreadsheet problem, and let the portal be the visible layer on top of a system that is actually true.