The 8 must-haves (TL;DR)
1. Work orders with multi-channel intake (web, mobile, email, QR, API)
2. Preventive maintenance scheduling (time, meter, condition triggers)
3. Assets in a parent-child hierarchy with history, costs, warranty, docs
4. Parts & inventory with multi-location, reorder points, barcode scanning
5. Mobile app with real offline capability and barcode/photo/voice support
6. Analytics for MTTR, MTBF, PM compliance, backlog aging
7. AI assistance grounded in your data — not a pooled chatbot
8. Security: SOC 2 Type 2, SSO/SAML on Enterprise, RBAC, data residency options
If a vendor cannot demonstrate all 8 on real data in one session, keep shopping.
1. Define the problem before you shortlist
The single biggest mistake in CMMS buying is shortlisting before defining the problem. Spend one meeting answering:
- How many users, sites, and assets will you cover in year one? In year three?
- What's broken today that forced this conversation? Backlog? Compliance? Downtime? Visibility?
- Who are the users? Field-first technicians, planners, executives — each has different mobile vs desktop needs.
- What systems does it need to talk to? ERP, HRIS, SCADA/IoT, finance?
- What budget band are you working within? Transparent plans vs quote-only?
Write the answers down before you book your first demo. Every demo turns into a checklist against this document.
2. Must-have modules
Work order management
Expect intake from web, mobile, email, QR, and API. SLA tracking. Priority levels. Attachments (photos, video, voice). Digital signatures. Full audit trail. Routing rules by asset, priority, or location.
Preventive maintenance
Three scheduling modes: time-based, meter-based, condition-based. Auto-generation of work orders on schedule. PM compliance tracking per asset and per asset class. Calendar visualization that shows load week by week.
Asset management
Parent-child hierarchy — organization → site → floor → system → equipment → component. Criticality scoring. QR, barcode, and NFC identification. Maintenance and cost history per asset. MTBF, MTTR, and total cost of ownership tracking.
Parts & inventory
Multi-location stock. Automated reorder points with lead-time awareness. Barcode and QR scanning. Issue-to-work-order tracking. Cycle counting.
Mobile
True offline: create, complete, scan, and photograph from anywhere. Conflict resolution on sync. Fast enough for a 3-year-old Android phone. Barcode and QR scanning from the camera. Voice-to-text for notes.
Reporting & analytics
Executive, operations, and asset dashboards. MTTR / MTBF / PM compliance / backlog aging. Custom reports with filter and export. Integration with BI (Power BI, Tableau, Looker) via API or direct connector.
3. Mobile offline matters more than demos suggest
The most common post-purchase surprise is that the "mobile app" works fine in a demo but falls apart in a basement, mechanical room, or on a plant floor without signal.
Ask to see:
- Opening and completing a work order with the device in airplane mode
- Scanning a barcode offline
- Capturing a photo offline and syncing later
- What happens on the second device when two technicians edit the same WO offline
If the vendor declines to show these, run.
4. AI questions to ask
AI-powered CMMS features in 2026 range from "glorified search" to "agentic assistants that draft and route work." Ask:
1. Is the AI grounded in my tenant data, or a pooled model?
2. Does it just generate text, or does it call CMMS APIs (tool use)?
3. Is it available on mobile and in my language?
4. How is its output logged in the work-order audit trail?
5. What happens if I disable it — do core features still work?
For a reference point, LAYLA is FacilityLane's in-product AI assistant on the Enterprise plan.
5. Security and compliance
Non-negotiables:
- SOC 2 Type 2 report (not just Type 1, not "in progress")
- SSO/SAML on Enterprise plans
- Role-based access control with scope by site and asset class
- Data residency options if you operate in the EU or UAE
- Multi-tenant isolation — ask how. "Logical separation" is not the same as database-level RLS.
- Audit logs that meet your compliance regime (healthcare, aviation, food, defense)
6. Pricing models and traps
Transparent published pricing matters more than the headline rate. Common traps:
- "Call for pricing" on every tier — expect prices that scale with urgency, not value
- Features locked to the top tier that should be in the middle (SSO, API access, reporting)
- User-count tricks: per-named-user vs per-active-user vs technician-only vs read-only differences can swing TCO by 2–3x
- Implementation fees out of proportion to the deal size
- Yearly commitments that look cheap but include a 20%+ annual escalator
FacilityLane's public pricing: Free ($0 forever), Starter ($15/user/mo), Professional ($49/user/mo), Enterprise (custom). See pricing.
7. Integration must-haves
Depending on your stack:
- ERP (SAP S/4HANA, Oracle NetSuite, Dynamics 365, Infor, QuickBooks, Xero) for parts/PO/invoices
- Fleet & IoT (Samsara, Motive, MachineMetrics, Ignition) for meter data
- Communication (Slack, Teams, Email via SendGrid, SMS via Twilio)
- Analytics (Power BI, Tableau, Looker) for executive reporting
- Automation (REST API, Zapier for 7,000+ apps, webhooks)
Any vendor without a documented REST API is a dead end in year two.
8. A 30-day pilot plan
Buy on evidence, not demos. A rigorous 30-day pilot answers the real question: will this work for our team?
Week 1 — Setup
- Import 1 site, 50 assets, 10 users (planner, supervisor, 6 technicians, 2 read-only)
- Configure 1 asset hierarchy, 5 PM schedules, 1 inventory location
Week 2 — Field run
- Technicians work live WOs on mobile
- Test offline sync in the worst-signal location
- Create one service-request intake path (QR + email)
Week 3 — Reporting
- Run MTTR, MTBF, PM compliance, backlog aging reports
- Export to your BI tool
Week 4 — Integrations & AI
- Connect one ERP or Zapier flow
- Run AI assistant on real prompts — measure useful output rate
- Decision meeting
If the vendor won't support this pilot shape, you have your answer.