Digital Signage KPIs and SLA

Five, and uptime outranks the rest. Uptime — the percentage of time screens show the correct content, target 99.5% or higher. Content freshness — the share of screens showing content older than your threshold, measured from the last sync timestamp per player. Mean time to repair — how long from a screen failing to it being back on air, split by tier: diagnosis in 10–30 minutes, remote fix in 1–4 hours, hardware swap or firmware patch in 24–48 hours. Hardware failure rate — under 1% per year, tracked through RMAs and fleet health reports. And playback evidence — an audit trail of what actually ran on each player, which matters most when the screens carry paid advertising. Define all five before procurement, because they are also the clauses you will want in the SLA.

99.5% is the floor for anything an audience sees. That allowance is smaller than it sounds: 0.5% of a year is roughly 44 hours of wrong or missing content. SpinetiX hardware contributes 99.96%+ availability, from a 0.4% hardware failure rate measured over a ten-year window across thousands of deployed units — a fleet of a thousand players is expected to lose four to hardware failure in ten years. The remaining risk sits in the network, the CMS and human error, which is where most real outages actually originate. Alert on any screen offline longer than 15 minutes.

Add every recurring cost to the purchase price, then divide by screens and by years. Total cost of ownership = hardware + software licences + installation + electricity + network + maintenance + support + replacement cycles, over a 5–10 year lifecycle. The items that decide the outcome are rarely the ones on the quote: a budget Android player replaced three times in ten years, generating around five times the IT tickets, drawing 65 W where an HMP400 draws 9 W typical, ends up costing more than the SpinetiX player that ran the whole period without being replaced. Quarterly firmware patches instead of continuous OS maintenance is a labour line too, not just a security one.

Four numbers, all in writing. Hardware: 3–5 year warranty minimum with RMA inside 48 hours. Cloud CMS: a 99.9% uptime SLA with defined RPO and RTO. Local support: a committed response time — Media La Vista commits to 10 minutes for Partner Club members, with Tier 1–3 engineering located in the Middle East rather than a European escalation queue. And consequences: an SLA with no penalty for missing the uptime target is a marketing document, not a contract. Response time, resolution time and escalation path belong in the agreement, not in the proposal deck.

Fleet monitoring with an alert threshold you set. Arya Cloud checks player status continuously, and the dashboard shows per-player online/offline state, last sync time, firmware version, storage usage, network address, and a live screenshot of what is actually on the screen — so content accuracy is verified from a desk instead of from the site. Configure an alert for any player offline longer than a threshold, typically 15 minutes, and set it tighter on mission-critical screens. For estates already run from a NOC, SpinetiX exposes SNMP v2c (read-only, disabled by default) to Nagios, Zabbix, PRTG or SolarWinds, and SNMP traps give near-instant notification. The screenshot matters as much as the status flag: a player can be perfectly online and still be showing the wrong thing.

Yes, but the evidence comes from the management layer, not the player alone. Proof-of-play is a verified record of which asset played, on which player, at what time, for how long, and whether the play completed. SpinetiX players expose the interfaces that carry health and playback data — Status API, syslog and SNMP — while turning that into a reportable, tamper-evident record is the CMS's job. 123CMS, the customer-branded management layer Media La Vista builds on top of the SpinetiX platform, records per-asset playback events from every player it manages and writes them to an immutable audit log on infrastructure the operator owns. If your KPI set includes advertiser reconciliation or compliance reporting, specify that layer explicitly rather than assuming a player ships it.

Choose the metric by use case, and baseline it before deployment rather than after. Retail: sales uplift on promoted lines near the screens, promotional conversion rate, dwell time. Corporate: employee awareness survey scores, meeting-room utilisation. Wayfinding: the drop in reception-desk enquiries and visitor satisfaction scores. Healthcare: perceived wait time and information recall. The honest limitation: no signage platform, SpinetiX included, measures audience attention by itself — a player knows what it played, not who watched it. Dwell time and footfall need a separate third-party layer of camera analytics or sensor counters, with their own privacy and consent obligations, and sales lift comes from the retailer's own point-of-sale data correlated against playback logs. Playback evidence is something the vendor owes you; audience measurement is a project you fund separately.