How to Run a 30-Day OEE Software POC: Metrics, Success Criteria, and Go/No-Go Framework

A POC is not a live demo. It is a structured experiment designed to answer specific business questions with data from your own production environment, before committing to a multi-year contract.

See the OEE calculation this POC evaluates.

The Four Questions Your OEE POC Must Answer

• Does the system generate accurate OEE data for our specific machines?
• Does the data quality reveal improvement opportunities our current monitoring misses?
• Can operators and maintenance technicians use it without friction?
• Does the vendor's implementation support meet our expectations?

Scope Rule: Start With 3-5 Machines

Cover your range of equipment types, enough to validate connectivity and data quality, small enough to implement thoroughly in 30 days. The most common POC mistake is over-scoping at the start.

• Sensor installation / PLC connectivity on POC machines
• OEE platform configured with shift schedules, targets, and downtime reason codes
• User accounts created for operators and maintenance team on POC lines
• Day 7 checkpoint: OEE monitoring live and validated against actual production counts

Days 8-21: Operations and Learning

• Operators entering downtime reasons daily
• OEE data reviewed in morning production meetings
• Day 14 checkpoint: First OEE report, identify top 3 loss categories
• Days 15-21: Target one specific improvement from POC data and measure impact

Days 22-30: Evaluation

• Calculate OEE baseline for POC period vs historical estimate
• Document connectivity quality, data accuracy, and adoption observations
• Quantify financial value of OEE visibility gained during POC
• Day 30: Go/no-go decision meeting with evidence from your own operation

Define success criteria in writing before the POC begins, not after, when bias toward a specific outcome can influence interpretation.

Recommended OEE POC Success Criteria

• OEE data accuracy: Within 5 percentage points of manual production records for the same period
• Connectivity reliability: Data capture above 95% of scheduled production time with no unexplained gaps exceeding 30 minutes
• Operator adoption: 80%+ of downtime events coded with reasons (vs "unknown") by end of week 3
• Insight generation: At least one improvement opportunity identified from POC data not visible from previous monitoring
• Vendor responsiveness: All implementation issues resolved within 48 hours or escalated with a documented resolution plan

Note: Vendors who push back on specific success criteria are revealing capability gaps worth investigating before commitment.