28 Oct 2026

How to read an in-room ad performance report: the five numbers that matter

A report lands in your inbox with eleven columns and two charts. The five numbers that actually decide whether in-room TV ads are working, the ones that flatter, what a good report refuses to show, and how to read a soft week without panicking.

You ran a pilot week. A report lands in your inbox with eleven columns and two charts. Most operators read the top line, see a small number, and either celebrate or panic. Both reactions are usually wrong. Here's what we put in our own reports, what each number actually measures, and which ones flatter.

Number one: impressions, with the denominator

An impression means the screen played the ad while a guest could plausibly have seen it. We count it when the playout fires on an active screen during an occupied night. The number alone is nearly meaningless. 40,000 impressions sounds big until you learn it came from 220 occupied room-nights across six rooms.

Ask for impressions per occupied room-night. That's the honest unit. A single room that plays a restrained ad load for a week should produce a few hundred impressions. If your report leads with a raw total, divide by occupancy before you react to it.

Number two: opportunity seconds, not spot length

Every report we produce lists the ad load per hour, in minutes and seconds, per daypart. Broadcast television in a living room runs roughly seven minutes of ads an hour. A bedroom in a short-stay property is a different room. Guests are dressing, packing, or half asleep, and tolerance is lower.

The number that matters is how many of your ad minutes fall inside the hours guests actually watch. A screen that plays four minutes an hour between 19:00 and 23:00 delivers more real attention than eight minutes an hour spread across a full day including dead afternoon hours. If the report shows a single blended figure, ask for the daypart split. If nobody can produce it, that's a finding.

Number three: screen-active rate

This is the boring number that quietly decides everything else. Screen-active rate is the share of occupied nights where the TV was on at all during ad playout windows. On a property with broken remotes, guest-log-out habits, HDMI-stick confusion or a router in the wrong room, this rate can sit at half of what it should.

When revenue underperforms a projection, the cause is more often a low screen-active rate than weak advertiser demand. The fix is hardware and setup, not pricing: check input defaults, fix pairing, make sure the TV reboots into the right input, and keep a working remote taped or tethered in the room. We covered the hardware side in detail in our guest-proofing piece.

Number four: attribution you can count

Every platform, us included, will show you "influenced revenue" or a view-through figure. Treat those as context, never as proof. The number that survives scrutiny is direct attribution: a guest scanned a QR or tapped a menu tile, and the order landed. That chain can be counted, logged and shown to a sceptic line by line.

A competent weekly report separates hard conversions from assisted ones and shows the raw event log behind each. If your report blends them into one optimistic column, push back. Advertisers renew on the hard numbers, and operators who quote soft numbers to local buyers tend to get found out on the second meeting anyway.

Number five: complaint mentions

The last number is the one most operators skip. Count the guest reviews and messages that mention the TV or anything that appeared on it. Across a healthy pilot this sits near zero. Two mentions in a week is a signal. Four is a settings problem.

The point of an in-room network is that it earns without costing the guest experience. Ad load settings can be dropped in minutes. A pilot that produces strong impression numbers and one scathing review about "ads on the holiday TV" has failed on the metric that matters most to a five-star rating, and no revenue line makes up for a lost review.

One bad week is not a verdict

Pilots wobble. A single soft week comes from check-in patterns, a rainy stretch, one room off the books for a repair, or an advertiser campaign that ended early. Read the trend: our pilot reports carry at least four weeks before anyone should draw a conclusion, and a stable or rising screen-active rate with a soft conversion week is usually a timing problem, not a platform problem.

Three signs a week was noise rather than signal: occupancy dropped by a fifth or more, the screen-active rate held steady while conversions dipped, or a single advertiser's campaign paused midweek. Three signs it was real: the screen-active rate fell, complaint mentions rose, or ad load drifted up after a manual settings change nobody logged.

What a good report refuses to show

Two things we won't put in your report, because nobody can honestly produce them for your property from a desk.

First, a guaranteed revenue figure. Anyone who quotes you one before your pilot has run is selling a spreadsheet, not a system. Occupancy, ad load, screen-active rate and advertiser mix all move the number, and they're all specific to your rooms.

Second, audience demographics per room. Nobody can tell you who was watching at 21:40 on Tuesday. We can tell you the screen was on and the ad played. Demographic claims in-room contexts come from modelled panel data, and their error bars don't survive contact with a 30-unit building.

Reading your first report

Start with screen-active rate, because it's the input to everything else. Then read impressions per occupied room-night, then the daypart ad load. Then hard conversions. Then complaint mentions, before you read anything else flattering. That order will keep you honest about what a pilot actually showed, and it's the same order we use when we decide whether a property is worth rolling to more rooms.

If a number in your report doesn't have a denominator, a daypart, or an event log behind it, it isn't a result yet. It's a claim. Ask us for the raw version. We would rather show you the log than lose a renewal.