Build or subscribe

Thousands of associations collect a membership fee. Thousands of hotels take bookings, and thousands of shops sell online. When many organisations share a need, a company has usually built a ready-made service for it. Doctolib, Vinted and Leboncoin are huge custom builds, because what they do is their whole business. A climbing club or a bakery has no reason to build its own.

A ready-made service is ready in days. You pay a subscription, a fixed price each month, or the service keeps a share of each payment. The provider handles security, payments and updates, for all its customers at once. HelloAsso, for example, handles memberships, ticketing and donations for French associations. It's free for them: the people who pay can add an optional tip, and the tips fund the service.

A custom build is a different scale. It costs thousands or tens of thousands of euros before it works, and takes weeks or months. Then it needs maintenance every year: hosting, updates, fixes, and someone who understands the code. A common estimate puts that maintenance at 15 to 20% of the build price, every year.

You're a junior consultant, and three clients want a clear answer with rough figures this week.

• A student association with 300 members wants members to join and pay their €20 fee online. A member's cousin offers to build a custom site for €2,000. Recommend HelloAsso. It does exactly this: the form, the payment, the receipts and the member list. It's free, and it's ready in an afternoon. The cousin's site would do less, and someone would have to look after it every year, long after the cousin has left.
• A family hotel with 15 rooms wants guests to book and pay on its own site, as well as on Booking.com. An agency quotes €18,000 for a custom booking system. Recommend a ready-made booking tool, for a monthly fee or a commission on each booking. The hardest part of hotel booking is keeping the free rooms the same on every site, so a room is never sold twice. These tools already do it, often with a channel manager. A custom system would have to solve it too, then cost a few thousand euros a year to maintain.
• A start-up whose business is matching families with home carers for elderly parents, with its own rules about who suits whom. An agency quotes €60,000. Here the matching is the core business, and no service does it their way, so building it may be right one day. But €60,000 before knowing whether families want the service is a big risk. Test the idea first with a form, a spreadsheet and a person who does the matching by hand. If it works, build the matching, and still use ready-made services for payments, email and calendars.

Even inside a custom site, use ready-made pieces. An AI agent builds your portfolio, but a form service can deliver its messages, and a hosting service puts it online.

Two beliefs push organisations towards custom builds. The first is "a custom site is better, because it fits exactly". It fits on the day it's delivered. After that, every change must be paid for, and the new features that a platform adds for all its customers never arrive. The second is "a subscription is money wasted, while a custom site is ours". A custom site keeps costing money too. It needs hosting, updates and someone who understands it, every year. Owning code is only an advantage when that code does something nobody sells.

So compare the full price of each route over the same years. For the custom build, that's the build plus its maintenance, year after year. For the service, that's its monthly fee or its share of each payment, for as many years.

Then check the service against the real needs before you recommend it:

• Its fees: per month, per user or per payment, and what they become as the organisation grows.
• What it can't do: read its help pages, or ask its support.
• Your data: can you download it in a format another tool can read, if you leave one day? This is a data export. Without one, you're locked in, and leaving becomes very costly.

An agency tells you: "The platform can't do that. You'd have to leave it for a custom site." That sentence compares one feature with a whole new site. Leaving costs far more than the feature. You pay for the custom site, then for its maintenance every year. And you lose everything the platform did for you: payments, security, updates, and often support.

So before anyone spends money, try the cheaper answers in order. Does the platform, or an add-on, already do it? Platforms have settings that few people know, and many have a store of add-ons, each for a monthly fee. Can the need be met another way, or dropped? Often the person asking cares about the result, not the exact method. Can just that piece be built and plugged into the platform?

• A climbing club wants its HelloAsso form to look exactly like its own site. It's only about looks, and nobody joins a club less because the payment form looks like HelloAsso. Keep the platform, add the club's logo if it allows it, and link to the form from the club's site.
• A bike shop on Shopify wants customers to design their own bike from 200 parts. Customers keep asking for it, so it sells bikes, and it's worth paying for. Look for a configurator add-on first. If none fits, have only the configurator built and plugged into the shop.
• A restaurant wants to refuse bookings from people who didn't come last time. Ask the booking tool's support first: many tools offer a card guarantee against no-shows. Otherwise, a short list of names, checked by hand, is enough.

Then choose one route, and say why: what the feature brings in sales, members or time, against what it costs. Count the add-ons' monthly fees too. A shop with ten paid add-ons can cost more each month than the platform itself.

Ask these questions whenever someone proposes to build something, or to leave a tool that works.

First, ask whether a ready-made service already does it. For memberships, bookings, shops, forms and ticketing, the answer is usually yes. Then ask what makes this organisation different from others. Build only that part, after testing the idea cheaply, and subscribe for the rest.

For a custom build, ask for its yearly cost after launch, not only its price. For a service, ask for its fees as the organisation grows, what it can't do, and how to get the data out. When an agency says "the platform can't do that", ask it to price the smaller routes too: the add-on, or only the missing piece. And search the platform's help pages before you conclude that something is impossible. Someone has usually asked before you.

Then give a clear recommendation, with rough figures for each route.