The ceiling on in-room screen revenue is set by occupancy and dwell time, which we covered in how much a Smart TV can actually earn per room. But the ceiling is not the number most operators miss. The number they miss is the conversion rate on what the screen actually offers, and that is entirely under your control. A unit with a mediocre offer menu at 90 percent occupancy earns less from upsells than a mid-occupancy unit with an offer list built around what guests already ask for.
Start from what the phone log already tells you
Every short-stay operator already owns the data for the best possible offer menu: the questions guests ask by WhatsApp on arrival day. Late checkout requests. Extra towels. Where to book a transfer. Early check-in because the flight lands at 6am. Breakfast options. Go through the last three months of guest messages and count the requests. Anything asked more than once a month belongs on the screen, priced, because the guests who message you are the visible fraction of the ones who would buy silently if the option was in front of them at check-in.
The corollary matters just as much: anything guests never ask about does not belong on the menu. A generic list of partner discounts and city tours copied from a hotel directory converts close to nothing and trains guests to ignore the screen.
Rules for an offer that converts
- Price it on the screen. An offer described but unpriced forces a conversation, and guests on holiday do not start conversations to spend money. Priced offers sell at several times the rate of "ask us" offers.
- Cap the menu at six to eight items. Beyond that, choice turns into wallpaper. Rotate seasonally instead of lengthening.
- Sell time, not things. Late checkout, early check-in, mid-stay cleaning and fast-track airport services are the four highest-converting upsell categories in short-stay inventory. They cost the operator little and guests perceive them as valuable.
- Limit purchase windows honestly. "Available until 9pm tonight" beats a standing offer. Deadlines are not a dark pattern when the constraint is real: the kitchen really does close, the transfer desk really needs three hours' notice.
Sequence matters more than content
Guest attention peaks twice: in the first hour after check-in, and the evening before checkout. Structure what you show around those windows. On arrival night, breakfast and wifi-friction items sell. Mid-stay, cleaning and consumable top-ups. In the final 24 hours, late checkout, luggage storage and the transfer back to the airport. A menu that shows the same list all week wastes both peaks.
The single most common mistake is the reverse: pushing checkout-day offers at check-in, when the guest has not yet thought about leaving, or worse, pushing them at 6am on departure morning when the decision window is already closed.
Measure it like a shelf, not a screen
Review the menu monthly with three numbers per offer: how many guests saw it, how many bought, and margin per purchase. Kill anything under one sale per 40 stays and replace it with the next item from your request log. This is standard retail thinking, and it is why operators with dull-looking but well-tended menus out-earn operators with beautiful looping content nobody can act on. If the platform you use cannot report views per offer, that is a platform limitation worth knowing before a pilot, alongside which smart TV platforms even allow third-party content in the first place.
A worked starter menu
- Late checkout to 2pm, priced, purchasable until 9pm the night before.
- Early check-in from 9am, subject to availability, shown only when the prior night is free.
- Mid-stay refresh clean for stays of five nights or more.
- Airport transfer, both directions, with a booking cutoff three hours out.
- Breakfast delivery window for the first morning only.
- Direct return-stay booking with the OTA commission split toward the guest.
Six items, every one traceable to a request guests already make. Get those live, run them for a month, and let the per-offer numbers tell you what the second month looks like.