Juniper Mist SLEs & Premium Analytics — User Minutes, Classifiers & 13-Month Dashboards

First — the wrong answer that gets engineers stuck

🧒 ELI5: Imagine your school takes attendance every minute for every student. At the end of the day it doesn't just say "school was fine" — it says "95% of student-minutes were good, and the bad 5% were all in Room 12 because the fan was loud." That per-minute report card is what Mist does for Wi-Fi.

🏛 Architect view: SLEs are Mist's experience-layer telemetry abstraction. They convert raw per-client, per-minute events into normalised success ratios with attributed root causes, so the experience layer (what the user feels) is decoupled from the infrastructure layer (what the AP reports). Premium Analytics is the long-horizon data lake on top — same events, 13-month retention, cross-stack joins.

A user complains: "Wi-Fi is slow today." The newcomer logs into the AP, sees green lights, checks CPU, finds nothing, and replies "looks fine on my end". The ticket bounces back angrier. This is the trap — they were checking whether the equipment is healthy, not whether the user's experience was good. Those are different questions.

Juniper Mist flips this. Instead of asking "is the AP up?", it asks "did each user get a good minute of service?". That single shift is the whole point of SLEs. You stop staring at device health and start measuring the human outcome. The complaint "Wi-Fi is slow" becomes a number you can argue with, slice, and fix.

① What an SLE actually is — the User Minute

Here is the unit everything is built on: the User Minute. Take one client, watch it for one minute, and decide: was that minute good or bad against the target? Do that for every client, every minute, all day. An SLE percentage is simply good user-minutes ÷ total user-minutes.

So a Coverage SLE of 96% means 96 of every 100 monitored client-minutes had a signal at or above your threshold. The missing 4% are the minutes that hurt — and Mist will tell you which clients, which APs, and which root cause owned them. This is why SLEs scale: a 5,000-user campus produces millions of user-minutes a day, and the maths stays the same.

The big picture — experience layer vs infrastructure layer

👉 So far: an SLE is good-minutes over total-minutes; there are 7 wireless SLEs; and the score is about experience, not equipment. Next: what each SLE actually watches.

The 7 wireless SLEs — what each one watches

② Classifiers — turning a dropped % into one root cause

A percentage tells you how much hurt. A classifier tells you why. Mist attributes each failed user-minute to exactly one classifier, and the classifiers for an SLE sum to its total failures. They appear as the coloured bars on the right of each SLE block.

Most classifiers then split into sub-classifiers for the exact mechanism. So you don't stop at "Successful Connects failed on Authorization" — you drill to "Authorization → EAP/802.1X timeout" and now you know to look at the RADIUS server, not the Wi-Fi.

The drill-down funnel

③ Thresholds & scope — make the score reflect YOUR network

A default SLE threshold is a starting point, not gospel. Coverage defaults to roughly -72 dBm and Time to Connect to about 2 seconds — but a hospital with medical telemetry may demand -67 dBm, while a warehouse may accept -75 dBm. You set the bar; Mist scores against it. Set it too loose and every SLE is green and useless. Set it too tight and you drown in false alarms.

The second half is scope. Every SLE can be read by site, AP, band, WLAN/SSID, client, OS, or time window. The same Coverage gap looks like "the building is bad" at site scope, but reveals "only AP-4F-12 on 5 GHz at lunchtime" once you slice it. Always narrow scope before you act.

Decision aid — is the SLE telling you to act?

④ Premium Analytics — when 30 days isn't enough

Standard Mist analytics keeps about 30 days of data — perfect for "what broke this week". Premium Analytics is the paid add-on that extends retention to 13 months or more and adds business-grade reporting. The question on every exam and every budget meeting is the same: when is it actually worth it?

Premium Analytics earns its cost in three situations: (1) you need long-horizon trends — month-over-month SLE, capacity planning, "prove the upgrade worked"; (2) you need cross-stack reports spanning Wireless & Location, Wired and WAN in one dashboard; and (3) you need business location analytics — occupancy, engagement and zone insights that turn the same Wi-Fi/BLE data into real-estate, staffing and retail decisions.

Standard analytics vs Premium Analytics

👉 So far: Premium = longer retention + cross-stack + business analytics. Now the part students under-rate — what occupancy and engagement actually mean.

Occupancy, engagement & zone — the business lens

This is where Wi-Fi stops being IT and becomes a business sensor. By analysing wireless devices, BLE tags and BLE-app devices, Premium Analytics produces:

• Occupancy — how many people are in a zone over time. Drives real-estate, capacity-compliance and cleaning/staffing decisions.
• Engagement — visits, dwell time and movement patterns across a floor. Retail uses it for product placement; campuses for space utilisation.
• Zone insights — the same metrics drilled to a named zone (a meeting room, a store aisle). A blue zone name on the dashboard cross-launches straight into that zone's detail.

Org > Analytics > Premium Analytics
  └─ Dashboard stack: [ Wireless & Location | Wired | WAN ]
       └─ Occupancy Analytics  → Zone Occupancy Insights
            zone "Cafeteria-GF"   peak: 312   avg dwell: 23m
            zone "Meeting-3F-B"   peak:  41%   avg dwell:  9m   (blue = cross-launch)

Retention window .......... 13 months (vs 30 days standard)
Data sources .............. Wi-Fi clients + BLE tags + BLE-app devices
Zone "Meeting-3F-B" ....... underused — peak 41%, dwell 9m
Recommendation ............ consolidate; SLE history confirms no coverage loss

One-glance cheat-sheet