Custom CMMS software: when to build your own maintenance system

Build a custom CMMS when your maintenance depends on data that lives elsewhere (production systems, ERP, machine controllers), when your audit trail requirements do not fit a standard product, or when your technicians will not use a generic tool. Otherwise, a well-configured off-the-shelf CMMS is cheaper and faster to roll out. The decision rests on adoption on the floor, integrations and total cost over several years.

Search results for this topic are dominated by CMMS vendors and software directories, each with a product to sell. This article does not rank any vendor. It helps you work out whether you need configuration, a custom build, or a hybrid of the two.

What a CMMS is, and what it is not

A computerized maintenance management system exists to look after your own assets: production machines, buildings, vehicles, technical installations. It answers four questions: what assets do we have, what work do they need and when, who did it, and what did it cost.

Two neighbouring tools are often confused with it:

• field service software, which dispatches technicians to customer sites with scheduling, quotes and invoicing. We cover it in custom field service software;
• production tracking (or MES), which follows work orders on the shop floor and machine stops. See production tracking software.

The three overlap: a machine stop logged by production should open a maintenance request in the CMMS. That handoff is often where standard tools fall short.

What off-the-shelf CMMS products do well

Established CMMS products are mature. Without any development, they cover:

• an asset register organized as a hierarchy (site, building, line, machine, component), with documents and QR code labels;
• maintenance requests that any employee can submit;
• work orders with priority, status, labour time and parts used;
• preventive maintenance triggered by calendar or by meter readings (running hours, mileage, cycles);
• spare parts inventory with reorder points;
• standard metrics such as mean time between failures and mean time to repair;
• a mobile app for technicians.

If your needs fit this list and your assets are conventional, start there. The general reasoning is laid out in build vs buy software.

Spreadsheet, off-the-shelf or custom

A spreadsheet is an honest first step. It becomes a liability when several people edit it at once, when preventive work depends on someone's memory, or when nobody logs anything from the field. Our guide to replacing Excel with a custom application covers that transition.

Four situations where custom is worth it

Maintenance depends on other systems

Meter-based preventive maintenance needs running hours or cycle counts. In a standard CMMS, someone reads the meter and types the number in. In a CMMS connected to machine monitoring or production tracking, the readings arrive on their own and the work order opens at the right moment. The same goes for parts: if purchasing runs through the ERP, the CMMS should read stock and raise purchase requests without anyone retyping them. Vendors sometimes offer these connectors, often as a paid extra, and rarely for a less common ERP or controller. Our approach is described in integrating a custom app with your ERP.

Your audit trail does not fit the boxes

Some organizations must prove, with records, that a check was done, by whom, when and with what result. Regulators set the rhythm. In the United States, the OSHA lockout/tagout standard, 29 CFR 1910.147 requires employers to inspect their energy control procedures at least annually and to certify each inspection with the machine, the date, the employees involved and the inspector. In Great Britain, regulation 9 of LOLER 1998 requires, for lifting equipment exposed to conditions that cause deterioration, a thorough examination at least every 6 months for equipment that lifts people and for lifting accessories, and at least every 12 months for other lifting equipment, unless an examination scheme sets different intervals.

An off-the-shelf CMMS can schedule these checks. It is less good at carrying your own rules: who may sign off, which evidence is mandatory before a work order can close, which report the auditor or insurer expects, and how to stop a signed record from being edited afterwards. When those rules pile up, a custom build starts to make sense.

Your asset base is unusual

A typical CMMS assumes a fixed plant. Other asset bases fit it poorly: a fleet of construction machinery that moves site every week, equipment you lease and install at customer premises, a property portfolio spread across many locations, laboratory or medical devices with their own calibration cycles. If the hierarchy the tool imposes does not match reality, every record becomes a workaround.

Technicians will not use the tool

This is the most common reason and the least admitted. A technician who needs five screens to close a ten-minute job will end up writing it in a notebook. A custom CMMS can be cut down to what your team actually needs: scan the label, see the machine's history, take a photo, close the job. The choice between web, installable and native apps is covered in web app vs PWA vs native app.

What a CMMS costs: subscription vs custom build

Off-the-shelf CMMS products are usually priced per user per month. As a public example, MaintainX listed on its pricing page, as of 7 November 2026, $20 per user per month for its Essential plan and $65 for Premium on annual billing, with Enterprise priced on request. Requesters (staff who only report problems) are free, and some integrations, such as a custom ERP connector, may cost extra. Open API access is listed from the Premium tier.

Run the numbers on your own case. Twenty-five technicians on a $65 plan come to $19,500 a year, before integrations, configuration and training, and every year after that. That is not an argument for building. It is the figure to put on the other side of the scale.

The price of a custom CMMS depends mostly on the number of integrations, offline support, compliance rules and data migration. We do not quote a range without knowing your scope; our article on custom business application cost explains what moves a quote. Then add the cost of maintaining the software itself, broken down line by line in software maintenance cost.

Two points change the comparison:

• you do not have to pay for development up front. With our subscription model, you pay a monthly fee that covers hosting, maintenance and enhancements, much like an off-the-shelf product;
• version one does not have to do everything. A CMMS focused on your critical assets costs far less than a full clone of a market product.

What version one should include

Custom CMMS projects go wrong when they try to match a mature product feature for feature. A good first version runs one complete cycle on a narrow scope:

1. the critical assets of one site, each with a QR code label;
2. the maintenance request, submitted by anyone in a couple of taps;
3. the work order, with photos, labour time and close-out;
4. calendar-based preventive maintenance on those assets;
5. a history visible from each asset's record.

Parts inventory, purchasing, meter-based schedules, advanced dashboards and sensors wait until that first loop is in daily use. AI-driven predictive maintenance, heavily marketed, needs years of clean history. It comes last, not first.

Where the real difficulty lies

• The asset hierarchy. It shapes everything: preventive plans, history, costs, metrics. Get it wrong early and it is expensive to rework. Design it with the maintenance manager, not only with management.
• Data migration. Asset history sits in spreadsheets, notebooks and the memory of senior technicians. Decide what is worth migrating; the rest can stay available as an archive. See legacy data migration.
• Offline mode. It must be designed from day one, including what happens when two technicians edit the same record without a connection.
• Permissions. Who can create, approve, close, or edit a signed report? In a regulated setting, an approved record is not edited, it is amended with a trace.

Common mistakes

• Starting from features instead of needs. A list of fifty modules says nothing about what your team will do on Monday morning.
• Designing without technicians. They do the data entry. If they skip it, no metric is reliable.
• Forgetting requesters. An operator reporting a breakdown should not need training or a complicated account.
• Promising predictive before preventive. Without reliable history, there is nothing to predict.
• Ignoring the exit. Code, data and documentation: rights are defined case by case in the contract. Read that clause before you sign.

The hybrid route

A sensible middle path keeps an off-the-shelf CMMS for what it does well and builds what is specific to you alongside it: a connector to your ERP or controllers, a compliance sign-off screen, a portal for maintenance subcontractors. First check that the vendor exposes an API on the plan you pay for (entry tiers do not always include it) and allows a full export of your data.

Where to start

1. List your critical assets: those whose downtime costs most or creates a safety risk.
2. Write down your obligations: statutory inspections, registers, reports your insurer or customers ask for.
3. Shadow a technician for a day: what they note, where, and what gets lost between the floor and the office.
4. Spot double entry between maintenance, production and purchasing.
5. See the tool before committing a budget.

For that last step, we build a free visual prototype and present it by video call within 5 working days: your technicians see the close-out screen or the asset record before anything is developed. The approach is explained in see your software before you pay.