On-call cover is the line item nobody quantifies. It is sold as availability, lived as standby, and the contract rarely says what time it starts. So everyone pictures something different: the owner thinks of the night, the manager thinks of the evening, the guest writes whenever the guest writes.
There is a simple way to settle it. Count the messages, hour by hour, over a stretch long enough not to be a single season.
- The curveWhere 461 guest messages really land
- The metricWhy answering faster is invisible
- The costFour designs, in hours per week
Where the load actually sits
We pulled 461 inbound guest messages from our database, received by a host with three listings over eight months, and sorted them by the local time of the listing. Cut into three hour blocks, they take a very clear shape.
Guest messages received, by three hour block
- 0 to 3
- 10 messages
- 3 to 6
- 0 messages
- 6 to 9
- 38 messages
- 9 to 12
- 81 messages
- 12 to 15
- 66 messages
- 15 to 18
- 64 messages
- 18 to 21
- 119 messages
- 21 to midnight
- 83 messages
One block in eight is empty. The busiest one carries 119.
461 guest messages received by a host with three listings, pulled from the database over eight months ·Two things stand out. The first is the empty block: across eight months, not one message between three and six in the morning. The second is that the peak sits in the evening, while a management business usually plans its working day around the afternoon.
Sliding the window instead of cutting it tightens the maximum further. The stretch from eight to eleven in the evening carries 153 of the 461 messages, a third of the total inside an eighth of the day. That is the slot worth holding. It lasts three hours.
The whole night, midnight to six, carries 10. That does not make it risk free: a lock refusing a code at one in the morning is still an emergency. But an access emergency is not a message flow, and it is handled with a procedure and a phone number rather than a continuous presence at a screen. The full hour by hour curve is published in the article on common guest questions.
Response rate does not pay for the night
Here is the point management offers avoid, because it removes a selling argument. The Airbnb response rate is the percentage of new enquiries and booking requests you responded to within 24 hours over the last 30 days. The window is twenty-four hours. A message received at ten in the evening and handled over breakfast counts exactly like a reply in eight minutes.
In other words, intraday speed does not exist in that metric. Overnight cover does not lift the figure shown on the listing, nor the Superhost requirement, which asks you to respond to 90% of new messages and to accept or decline new booking requests within 24 hours.
Only one situation genuinely punishes inaction: your response rate will drop if you leave a booking request unhandled and it expires. An ignored request counts as no reply, whereas a reply in the twenty-third hour is still a reply. The operational instruction that follows fits on one line: handle inbound requests before anything else, including before questions from a stay in progress.
What the evening slot buys therefore sits elsewhere. It buys the experience of the guest arriving, the meal, the code that will not work, the question about the heating. None of that surfaces in a public metric, all of it surfaces in the reviews. Say it to the owner in those words, otherwise the line on the quote looks expensive for a number that never moves.
What each shape of cover costs
Cover is priced in hours per week before it is priced in money. Four common designs, put on the same scale.
Hours of cover per week, four designs
Illustrative model- 8 to 11 in the evening
- 21 hours
- 7 to midnight
- 35 hours
- 8 to midnight
- 112 hours
- round the clock
- 168 hours
The evening slot fits in 21 hours. Continuous cover asks for eight times more.
Each design held seven days a week, with no rest day and no rotation, cover counted as raw presence hoursSet against the measurement, the gap becomes a trade. The eight to eleven slot covers a third of the messages for 21 hours a week. Stretched from eight in the morning to midnight, it covers 95% for 112 hours. The remaining 5%, the night ones, cost 56 more hours every week. That ratio decides, not the comfort of a formula that promises everything.
The tipping point towards a team is documented elsewhere: it lands at the third listing, well before revenue pays for it. That is the subject of the guide on starting a management business.
Writing the cover into the contract
- Name a slot with two precise times, never a phrase such as broad availability. A slot is either held or not held, availability gets argued about afterwards.
- Separate the messaging slot from access standby. They share neither the hours, nor the trigger, nor the price.
- Set the pick-up time inside the slot, then the pick-up time outside it. Two numbers, not an adjective.
- Write the priority rule: a booking request comes before a stay question, because it is the only one whose expiry costs something.
- Plan for cover of the cover. A three hour slot held seven days a week by one person collapses in the first week of holiday.
Those five lines are worth more than a responsiveness pledge quoted to the minute. They describe what will be done, at what time, and by whom in case of absence. That is exactly the question the owner is asking without knowing how to phrase it.
- CopiloteSentioo drafts the reply from the facts of the listing, the person on standby reviews before sending.
- HybrideQuestions with a stable answer go out on their own during the slot, anything with a cost attached comes up.
- AutopiloteSentioo replies directly outside the slot, nights included. Anything beyond the covered scope waits for the person on standby.









