Five Reports Your Analytics Dashboard Should Show You

M ost captive portal dashboards are comprehensive and get ignored. Every possible number is on screen, so nobody builds a habit of looking at any of them. The fix is not fewer features. It is knowing which five answers you actually need, and making sure you can get to each of them in one click.
1. Usage over time
The shape of your network across a day is the baseline everything else is measured against. Hourly or daily curves immediately show you the opening and closing rushes, the quiet afternoon, and any day when something unusual happened. It is also the chart you will want when you are asked to justify buying more bandwidth.
Look at it with the zoom and filter controls rather than as a single aggregate number. The interesting question is rarely “how much bandwidth did we use in October” but “why did the 14th look like the 13th”.
2. Who is online right now
A live connection list is the operational view: every active session, the device, the profile it runs under, the access point it is attached to and how much it has consumed. Two things fall out of it immediately.
First, it is the tool for the incident. A guest complains that the connection is slow, and instead of guessing you can look: is the session actually running, is it attached to the access point you think it is, has it hit a limit and stalled, or is there one device saturating the uplink.
Second, it is the tool for the enforcement. If a session has exceeded its allowance, or a device needs to be disconnected — by policy, by a security event, or by a support request — you can act on it there and then, rather than at the next scheduled maintenance.
3. Which access points are busy
Access point statistics tell you about the building rather than about the internet. The top access points list answers a question with a very direct operational consequence: where should the next unit go, and which ones are underused?
A single saturated access point in a conference room while the lobby unit sits at 5% utilisation is a placement problem, not a bandwidth problem, and no amount of internet capacity will fix it. Seeing that pattern is the difference between buying the right hardware and buying the wrong one.
4. How each profile consumed its allowance
Per-profile statistics are where capacity planning becomes concrete. You configured a profile with a daily data cap; this report shows who hit it, who came close, and who barely used it at all.
A profile that consistently sits at 90% of its cap is a pricing problem: either the cap is wrong for what you sold, or you should be selling more of it. A profile that never reaches 10% is dead configuration. Both are discovered in a minute here and guessed at for months elsewhere.
5. The individual heavy users
Aggregates hide the problem. A network can look healthy on average while a single device downloads a terabyte overnight, and the symptom is that everyone else has a bad experience. The top users list names that device, and the live list can then act on it.
In most cases this is not malicious. It is one guest casting a screen to a TV over the guest network, or a laptop running a backup. But you cannot manage what you cannot see, and the report is what turns “the WiFi was slow last night” into a specific, fixable cause.
Make the answers exportable
All five reports become far more valuable the moment you can take them somewhere else. Export to CSV or image for a monthly management pack, for a shift handover, or as evidence when a customer disputes a bill. The alternative is a screenshot, which nobody trusts and nobody can compare.
And check them against reality
One caveat worth stating plainly: analytics describe what the portal saw. If the portal is not seeing all of your traffic — because a segment of the network bypasses it, or because users are being authenticated somewhere else — then the dashboard is accurate and still incomplete. Review the coverage against the network you actually run, not the one you designed.
Used this way the dashboard stops being a report and becomes a control panel: five numbers that tell you whether the network is healthy, where the problem is, and what to change next.

No part of this document may be reproduced without prior written permission of WifiGem.