Fix a wrong People Counter result by reconciling the day’s Entered and Exited totals with real movement, because Current people is their net for that local date.
People Counter is a manual browser tool rather than an automatic detection system. It does not use a camera, sensor, badge reader, location feed, or AI model to observe arrivals and departures. Instead, you choose when to press Record entry or Record exit. A result can therefore look wrong because someone missed a movement, pressed the wrong control, corrected an earlier day on a later day, or opened the tool when browser storage was unavailable.
Open the private, browser-based People Counter and begin with the totals for the relevant date. Do not try to make the display look right by pressing the opposite button: unless that movement really happened, doing so would add a false entry or exit. The supported recovery process is to identify the type of error, reconcile genuine omissions, and preserve or rebuild the stored record only when necessary.
Identify which part of the result looks wrong
Start by checking the device’s local calendar date. Each interaction is attached to the local date on which it is recorded, and the visible current count starts at zero when a new local day begins. Earlier nonzero dates remain available in Daily history. This date behavior can make a sudden reset look like a counting error when the local day has simply changed.
For the selected date, read Entered, Exited, and Current people together. The relationship is:
Current people = Entered − Exited.
For example, 3 entered − 2 exited = 1 current person. Use the formula to understand the direction of the result, not to replace missing observations with guesses. The actual movement record still needs to confirm that each recorded action belongs in that total.
| What you see | Likely cause to check | Supported next step |
|---|---|---|
| The current count is higher than expected | An entry may have been recorded twice, or a departure may have been missed. | Compare Entered and Exited with the actual movements. Record a genuine missed exit once. An individually false entry cannot be removed in place. |
| The current count is lower than expected | An arrival may have been missed, or an extra exit may have been recorded. | Record a genuine missed entry once. Do not press entry to cancel a false exit unless an arrival actually occurred. |
| The visible count unexpectedly returns to zero | A new local calendar day may have begun. | Check Daily history for the previous local date rather than assuming the earlier record disappeared. |
| The count appears below zero | This is not a normal People Counter result. | Exit is disabled at zero, and the underlying logic rejects an exit that would create a negative net count. |
| The capacity alert appears unexpectedly | The current net count has reached or exceeded the whole-number threshold you entered. | Confirm your threshold setting. The alert is a reminder and does not change the count. |
| A record is missing after refreshing | Browser storage may be unavailable, unless you deliberately cleared the data. | Check the page’s storage notice. When storage is unavailable, an interaction can appear for the current session but cannot be retained. |
Keep the correction on the correct date in mind. If you discover a missing movement today but press the correction button today, the interaction belongs to today’s local date. People Counter does not provide a way to assign a new interaction to an earlier date. For a historical discrepancy, preserve the saved row first and keep any corrected working version outside the tool.
Correct a wrong People Counter result step by step
- Confirm the active local date. Open People Counter and check the device’s calendar date before pressing anything. A count made close to midnight stays on the local date when the interaction occurs rather than shifting to a UTC date.
- Review the relevant daily totals. Check Entered, Exited, and Current people for the active date. For an earlier date, open Daily history and identify its entered total, exited total, and net total.
- Compare the totals with a list of real movements. Use your own notes, a door log, or a staff member’s account to look for an omitted arrival, an omitted departure, a duplicate action, or a button pressed for the wrong direction. People Counter cannot perform this comparison automatically.
- Reconcile a genuinely missed movement. If someone was added to the occupied group but no entry was recorded, press Record entry once. If someone left but no exit was recorded, press Record exit once. The adjustment appears on the local date when you make it.
- Handle a false tap without inventing movement. The verified workflow does not include an individual undo, an editable totals field, or a cancellation button. Do not press the opposite control merely to balance the number. Export the record before taking destructive recovery action.
- Export the existing record. Choose PNG for a visual summary or CSV for row-by-row data. CSV contains Date, Entered, Exited, and Net people, with one row for every saved local date. Both exports are produced in the browser, and neither count nor export file is uploaded for processing.
- Clear and rebuild only when a false tap makes the in-tool history unusable. Clearing the tool’s saved browser data removes its history rather than deleting only the incorrect day. After clearing, begin a fresh manual record and re-record actual movement for the active local day. The verified workflow does not include an import step, so an exported CSV cannot be loaded back into People Counter.
- Verify the recovered record. Review the three totals again, confirm that the browser can retain the record, and export a fresh copy if the result will be used later. If you rely on the capacity alert, enter the intended whole-number threshold and remember that it is only an operational reminder.
Know which corrections are not possible in place
The available correction methods are intentionally limited. The verified People Counter workflow supports recording genuine entries and exits, reviewing daily history, exporting PNG or CSV, and clearing local browser data. It does not include per-event editing, a “cancel last press” action, historical backdating, or CSV re-import. That distinction matters because a display change alone cannot reveal which individual press was accidental.
The tool also cannot diagnose the physical event. It has no awareness of who crossed a doorway, so it cannot determine that a person really left merely because that would make the number look more convenient. If the source record is uncertain, preserve the exported totals and document the uncertainty rather than manipulating the counter to force a preferred result.
The capacity threshold does not provide an independent check. It alerts when the current net count reaches or exceeds the number you selected; it does not calculate occupancy from room size, identify people, certify safety, determine a legal limit, enforce access control, or replace trained staff and site procedures. Treat a reconciled count as a manual operational record, not a sensor-verified or code-compliant occupancy result.
Protect the corrected record before clearing browser data
People Counter stores its state under the tool’s local browser key. It is not posted to a server, connected to a sign-in, synchronized to another device, or maintained as an attendance database. Another browser begins with an independent empty record, so opening a second browser is not a way to continue the same count. If storage is unavailable, the page can still display an interaction in the current session while reporting that the record cannot be retained.
Export before clearing data or changing devices if the daily totals matter. A PNG provides a summary image with a total and recent date rows; it contains no URL and omits the optional custom label. CSV provides the structured daily fields and can be opened in a spreadsheet as an external working copy. Because these exports contain daily totals rather than an event-level log, they cannot identify which individual button press was wrong, but they preserve the figures that were stored before a reset.
For a repeatable procedure that combines the counter, daily checks, and retained files, see how to document People Counter use. Keep any externally edited spreadsheet clearly labeled as a working copy, because People Counter has no import path for sending it back into the local history.
Make the next count easier to verify
A short operating routine reduces mismatches. Designate one person to press the controls when practical, record each real movement once, and avoid using Record entry or Record exit as a test. Review Entered, Exited, and Current people at a consistent point in the local day, then compare them with your notes before exporting the day.
Set a capacity threshold only when it supports a clear internal reminder. Near midnight, include the local date in any written reconciliation so an interaction is not attributed to the wrong day. If browser storage is unavailable, treat the visible total as session-only and obtain a retained copy through export before the session ends.
For a pre-session review of common recording problems, see how to avoid common mistakes when using People Counter. The key habit is not to force the total with compensating taps: confirm real movement first, correct genuine omissions with the matching button, and rebuild the record only after preserving what you need.