RTLS vendor selection checklist — questions a demo will not answer.
Most RTLS vendor selection decisions ride on a polished demo. These are the questions that separate fit from theatre — environment, TCO, integration and what fails after go-live.
Accuracy & performance
- Quote accuracy as a distribution (e.g. 90th percentile), not a single best-case figure
- Show a reference doing your exact use case in a similar RF environment
- Commit to a paid pilot in your space — metal, racking and forklifts wreck lab numbers
Integration & data
- How does it integrate with our WMS/ERP/EMR/MES — with a named reference?
- Can we export all our data, and who owns it if we leave?
- What is the API, update rate and event model?
Commercials & lock-in
- Total cost over 3 years: hardware, tags, licences, install, support, batteries
- What is per-tag/per-site licensing, and how does it scale?
- What is the exit path — proprietary tags/anchors or open ecosystem?
Delivery & support
- Who does the site survey, and is it included?
- What are SLAs, firmware support lifecycles and spares availability?
- What does a realistic rollout timeline and resourcing look like?
Vendor-neutral selection checklist for locating programmes
Shortlist locating vendors only after the use case, accuracy band and integration target are written down. Otherwise every demo will look impressive and none will be comparable. Score each vendor against your environment and KPI — not against their lab brochure.
Accuracy and RF performance. Ask for 50th and 95th percentile results from sites like yours (metal, racking, clinical partitions). Require a production-load pilot plan. Probe multipath/NLOS handling, update rate at your tag population, and what happens when a locator/anchor fails. Reject accuracy claims that never state percentile or conditions.
Integration and data. Native connectors or open APIs to your WMS/MES/EMR/CMMS; event semantics you can own; export without ransom; latency budgets; and identity models that match your asset master. A beautiful map that cannot emit a reliable event is not a locating platform for enterprise use.
Commercials and lock-in. Five-year TCO including tags, batteries, licences, spares and professional services; price steps as tag counts grow; firmware support life; whether tags are multi-source; and exit assistance. Prefer standards (FiRa/UWB, BLE direction finding, EPC Gen2, GS1) over proprietary-only stacks unless the KPI truly demands them.
Delivery and support. References at your scale and industry; realistic rollout resourcing; local install partners; SLA definitions that match operational hours; and clarity on who owns RF redesign after layout changes. Ask what failed on a past project and what changed — vendors who cannot answer are selling aspiration.
Process tip. Same scripted scenario for every shortlisted vendor; independent scoring; no reseller in the evaluation chair. Use a vendor-neutral advisor when internal teams disagree or when density quotes diverge by more than ~30%.
Scripted scenario and scoring weights for a fair locating bake-off
Give every shortlisted vendor the same scripted scenario: defined floor area, tag counts, peak update needs, metal/clinical constraints, and one required integration. Score accuracy evidence, density rationale, integration readiness, TCO transparency, support model and exit options with published weights. Ban custom demo theatre that cannot be reproduced on your site.
Keep minutes of unanswered questions. A vendor that dodges percentile definitions or lock-in terms is telling you how the relationship will feel after signature. Where scores are close, prefer the architecture that preserves hybrid options and open event export over the flashiest map UI.
Exit, spares and firmware questions that belong in every RFP
Add explicit RFP questions: How do we export historical and live events if we leave? What is the spare-parts lead time for locators/anchors? How long is firmware supported on hardware bought this year? What is the process and cost to re-commission after a layout change? Who is on-call during our operational hours? Vendors that answer crisply are usually easier to live with than vendors with prettier demo apps and vague lifecycle terms.
Weight exit and lifecycle high enough that they can change the ranking. Locating infrastructure lasts longer than the average project team’s memory.
Bake-off facilitation without vendor capture
Facilitate scripted bake-offs with an independent chair, identical scenarios and published weights. Ban side demos that cannot be reproduced on your floor. Record unanswered lifecycle and exit questions in the scorecard. When scores are close, prefer open event export and hybrid-friendly architecture over UI polish. That process is how locating procurement stays comparable and defensible.
Share the scorecard with finance and operations before BAFO so commercial pressure cannot silently override accuracy and integration evidence.
Field note
Require written answers on accuracy percentiles, five-year TCO, event export and firmware life before BAFO.
If those four answers are vague, remove the vendor from the shortlist regardless of demo quality.
Document the ranking rationale so procurement, operations and IT can defend the locating choice after signature.
Frequently asked questions
What's the most common selection mistake?
Choosing on demo polish and a single accuracy number, with no pilot in the real environment and no look at 3-year TCO or lock-in.
Should we run an RFP?
Often yes — but only after requirements are written well enough to get comparable bids. We write those RFPs with you.
Do you sit on the vendor side?
Never for a client selection. We represent you, with no reseller margin and no platform of our own.
Can you shortlist vendors for us?
Yes — see our independent vendor comparisons and book a selection workshop.
Last updated:
Advisory note
This guide is written for operators selecting and deploying location technology — not for selling a radio. Use it to frame questions for vendors and integrators; bring TRACIO in when you want those answers scored independently against your site.