How to Read Your Attendance Report

Your count is done and the PDF is in your inbox. Now the real work starts: knowing how to read your Attendance Report so you can pull the right number for a sponsor, a grant annex, or next year's exhibitor pitch — and, just as importantly, so you understand what the report is not claiming. A good report is honest about its own limits, which is exactly what makes it believable. This is a section-by-section tour of what every part means and how to use it.

Key takeaways

• The headline total is one entries figure; the occupancy curve tells the more useful story of when your event actually peaked.
• Error bars and the confidence tier are features, not fine print — they are what make the number citable.
• The categories and direction split show who crossed and which way; dense-crowd intervals are flagged, never hidden.
• The QR verification page lets a sponsor confirm the number independently — that is the difference between a claim and evidence.

The cover and the headline number

Page one gives you the event title, the period counted, and the total attendance — the number of entries the engine counted across all your gates. This is the figure that goes in the press release. But read it alongside two things right next to it: the error range and the confidence tier. Together they tell you how hard the number can be pushed. A total of 18,400 at high confidence with a tight band is a figure you can defend in a renewal meeting. The report never hands you a bare number pretending to be exact, because that would be dishonest.

Entries, exits, and the occupancy curve

Because GateProof counts every crossing in both directions, your report separates entries from exits — and from those two it builds the occupancy curve: how many people were inside your event at each point in time. This is usually the most valuable page. A single total says "20,000 came." The occupancy curve says "you crossed 9,000 on site at 6:40pm, held above 7,000 for three hours, and drained by 9." That is the story a headline sponsor pays for and a safety officer needs. If you set your counting lines and in/out directions correctly, this curve is clean.

The hourly profile

Next comes the hourly (and finer, five-minute) breakdown — a bar chart of crossings through the day. Use it to prove when your audience showed up: the lunchtime family surge at a market, the pre-headline rush at a festival, the steady evening flow at a street fair. Sponsors buy specific slots; this page tells them exactly how many people their activation was in front of, hour by hour.

The direction split and categories

The direction split shows the balance of in versus out over time — useful for spotting a gate that mostly drained while another filled. Where the engine can distinguish categories, you will also see a breakdown (for example, cycles crossing a shared entrance). Two honesty notes here: some categories are marked beta because they are harder to detect reliably, and anything the engine is unsure about is labelled rather than guessed. If a category is not dependable for your footage, the report says so instead of inventing a split.

Error bars and confidence — read these properly

This is the part most people skim and shouldn't. The error bars express the range the true count likely sits in. Crucially, they widen where the count is genuinely harder — dense crowds, a congested gate, a dark clip — and the report flags those intervals explicitly. That is the opposite of a competitor's "99.5% accurate" banner with no evidence behind it.

The confidence tier is separate and just as important. It reflects how representative your footage was. Count a full event day end to end and you earn high confidence. Hand over a ten-minute spot clip and the report marks the figure indicative — it will not project a confident full-day or monthly number from a fragment. If you relied on partial footage, read how sampling turns a few hours into a full-event estimate to see how the projection is bounded. For the underlying maths of error and confidence, see what "accurate" really means.

The 60-second overlay clip

Every report links to a short annotated video: boxes on each person, fading trails, and a live counter ticking up as people cross your line. This is your sanity check and your proof at once. Watch ten seconds and you can see the count is real, the line is in the right place, and nobody is being double-counted. It is also the single most persuasive thing to drop into a sponsor deck.

The QR verification page

At the bottom of the cover sits a QR code and a public link. Scan it and it opens a verification page — the event, the period, the totals, the method, and the model version — served independently of you. This is what turns your report from "a number we produced" into "a number you can check." When a grant officer or a sponsor's finance team asks how do you know, you hand them the link. We cover how to stress-test the whole thing in how to audit a festival attendance claim.

Methodology, QC, and limitations

The final pages state the method plainly, list any quality flags on your footage (low light, camera shake, subjects too small), and spell out the limitations. A report that hides its weak spots is not trustworthy; one that names them is. The privacy footer restates the rule the whole product runs on: we count silhouettes, not people — no faces, no identities.

How to read your Attendance Report in practice

Reading the report well means matching the right page to the right audience: the total for the press, the occupancy curve for the headline sponsor, the hourly profile for activation partners, the verify link for anyone who doubts it. New to all this? Start with the pillar, how to count event attendance, then count your gates with a $199 Event Report at your next event to see your own report come to life. When you are ready to certify the full event, the pricing is fixed and simple.