Airbnb listing photo math: how many, in what order, and which one earns the click

In April 2025 I swapped the cover photo on a one-bedroom listing in the old quarter — same listing, same price, same five-star reviews. The old cover was a wide shot of the living room from the doorway, lights off, taken at 11am with the curtains half-drawn. The new cover was the same room at 6pm, lights warm, two glasses on the coffee table, shot from a corner crouched low so the ceiling beam led the eye toward the window. Nothing else on the listing changed. Bookings the next 30 days came in 38% above the prior 30-day rolling average. The photo at position 1 had been quietly losing me a booking every three days.

This post is the math on what that photo does, where each photo after it earns its keep, and the point on the count curve where adding more stops helping.

What the cover photo actually does

A guest searching "two-bedroom apartment, Lisbon, April 14–17" gets a grid of listing cards. Each card is one photo, one title line, one price, one star rating. The photo takes about 65% of the card's pixel area. The title takes 12%. Everything else — price, rating, badges — is the remaining 23%. The eye lands on the photo first and either commits to a click or moves to the next card. There is no second moment of attention. Either the cover earned the click or it did not.

Airbnb does not publish a click-through-rate breakdown of cover photos, but two pieces of evidence point to the same range. First, hosts who systematically A/B their cover photo (rotate between two finalists weekly and compare bookings) consistently report 30–60% swings in monthly bookings from cover changes with no other variable moved. That is a wider band than any other on-listing edit. Second, Airbnb's internal "Smart Pricing" and "search rank" both consume historical click-through rate as an input — and click-through rate is, on a results grid, the cover-photo number.

If you fix one thing this quarter, fix the cover. Everything else on the listing is competing for attention you have not earned yet.

What a working cover photo looks like

The covers that out-perform have four traits in common, in order of how much each one matters:

1. It is the room a guest most wants to be in. For a city apartment that is the living room nine times out of ten — not the bedroom, not the kitchen, not the entry hall. For a beach house it is the deck or the view. The cover answers "what does staying here feel like" in one frame.
2. It is shot at the time of day the room looks best. Most apartments photograph best between 5 and 7pm in summer, 3 and 5pm in winter. The window light is warm but not flat, and lamp light layers in. A bright-noon shot looks clinical; a 9pm shot looks gloomy. Pick the hour.
3. It has depth. A cover shot from a doorway looking across the long axis of the room reads as larger than the same room shot from the centre. Crouch slightly — 130 to 150 cm from the floor — so the ceiling shows. A shoulder-height shot makes ceilings look low.
4. It has one human prop. A cup on a side table, a paperback on the arm of the sofa, slippers by the door. One — not three. The prop tells the brain "people live here"; three props read as a stage set.

What does not move the cover photo's CTR: the resolution above 2000 px, the exact colour temperature, the watermark you put in the corner, whether you HDR'd the windows. Resolution past 2000 px is invisible on the search card. The watermark looks unprofessional and may be auto-flagged.

The diminishing-returns curve above 25 photos

Across every photo audit I have run, the booking-rate curve flattens between 20 and 30 photos. The biggest jump is from 5 to 15. The second biggest is from 15 to 22. From 22 to 30 is barely measurable. Above 30 is negative — the carousel becomes a chore and guests close the listing before reaching position 25.

The pattern, approximately:

The shape repeats across listing types — studios, family apartments, beach houses — though the slope above 20 is steeper for listings with strong outdoor features and gentler for compact urban units where there is genuinely less to shoot.

The reason for the flattening is that a guest scrolls until they have answered three questions: does the bedroom look like the cover, is the bathroom acceptable, and is there a working kitchen. That decision usually closes by photo 12 to 15. Photos 16 to 25 are tie-breakers — the closet, the balcony, the morning view. Past 25, the guest is bored and either books or leaves.

Aim for 22 photos. If you genuinely have more material that matters, push to 25. Past 25 you are adding weight without adding bookings.

Position math: where each photo earns its keep

Photos beyond the cover do not have equal weight. The carousel renders photos 1 through 4 in a 2×2 grid above the fold on desktop, and as a horizontal swipe on mobile where photos 1, 2, and 3 are partially visible without scrolling. The first row earns the second click — "view all photos" — and the second-click rate is the strongest predictor of a request-to-book after the cover.

A working order on Airbnb's grid:

The single biggest re-ordering mistake is putting the bathroom in position 18. Guests scroll the first row, do not see a bathroom, and assume you are hiding it. Move it forward — even an ordinary bathroom in position 4 outperforms a beautiful one buried at 18.

The second mistake is two consecutive shots of the same room. Photos 2 and 3 should be different rooms. If you have a stunning bedroom and want to show two angles, separate them: bedroom angle 1 at position 2, kitchen at position 3, bedroom angle 2 at position 5.

What to actually shoot, in order

If you are setting up a shoot from scratch, work the list in this sequence. The total time is 90 to 120 minutes for a one-bedroom, 150 to 180 for a two-bedroom.

1. The cover candidate — best room, best hour, depth + prop. Shoot 6 variations and pick later.
2. The two other "hero" room shots — wide, lights on, curtains open, no clutter.
3. The bathroom — wide first, then a detail of the sink or the rainfall shower head.
4. The kitchen — wide first, then the coffee setup, then the dining table set for two.
5. The bedroom(s) — wide with the bed made, then a detail of the linen or the bedside lamp.
6. The workspace, if one exists — the desk with the chair pulled out and a laptop closed on it. The "I could work from here" shot.
7. The view from each window worth showing — shoot at the same hour as the cover so light is consistent.
8. The balcony or outdoor area — wide, then a detail of the chair or the plant.
9. The entry / hallway / building front — only if it earns its place. A grim concrete corridor hurts; a tiled vestibule helps.
10. The neighbourhood block — the café two doors down, the bakery, the metro entrance. Three to five shots, not more.
11. The amenity proof — washer, dryer, AC unit, fast Wi-Fi router with a router-speed sticker visible if you have a fibre line.
12. The floor plan, if you can get one. A hand-drawn plan in black ink on grid paper photographs better than most architectural plans.

That is 20 to 25 photos. You will throw out a third of them in the edit. What survives is the listing.

For the operational side of running the shoot — when to do it, who to hire, what it costs — see linen inventory for short-term rentals for the kind of detail prep that shows in photos.

Refresh cadence: when to reshoot

The biggest decay is invisible. Photos do not get worse; the listing slowly drifts away from what the photos show. The sofa you photographed in 2022 has a sag. The wall got a scrape from a suitcase. The new toaster on the counter is white, not chrome like the one in photo 7. Each tiny drift on its own is fine; the cumulative drift is a listing that looks slightly off, and the guest cannot say why.

A working cadence:

• Cover photo: every 12 months, or after any room repaint, or after any furniture swap in the room shown. Even if nothing else changed, re-shoot. The light changes by season; the props get tired.
• Bathroom photos: every 6 months. Bathrooms photograph badly the moment a grout line darkens, and grout darkens in months.
• Bedroom photos: every 12 months, or whenever you change linen. New linen photographs differently; old linen photographs as old.
• Detail and amenity photos: every 18–24 months. Lower drift rate.
• Neighbourhood photos: every 24 months. The café closes, the metro station gets refurbished — refresh.

After every reshoot, run the cover-photo A/B: keep the old cover live for 14 days, swap to the new cover for 14 days, compare 30-day booking counts. Keep whichever wins by more than 8%. Below 8% is noise.

A 3-listing photo audit

To make the math concrete, three listings I audited in spring 2025 and what shipped:

Listing A — one-bedroom, ground floor, courtyard view. Started at 13 photos, cover was a wide living-room shot at noon. Audit found: bathroom at position 14, no bedroom in the first row, two near-duplicate living-room angles back-to-back. Changes shipped: re-shot cover at 6pm with one prop, moved bathroom to position 4, added 6 detail and neighbourhood shots, dropped one of the duplicate living-room angles. New count: 19 photos. 30-day booking rate: +28%.

Listing B — two-bedroom apartment, seventh floor, partial sea view. Started at 31 photos, cover was the sea view from the balcony. Audit found: the sea view oversold the apartment (guests arrived expecting full sea view and reviewed accordingly), positions 2–10 were all balcony angles, the second bedroom was at position 23. Changes shipped: swapped cover to the main living room at sunset (sea visible through window, secondary not primary), moved the second bedroom to position 5, cut the photo count from 31 to 22 by dropping balcony duplicates. 30-day booking rate: +14%. Review rating after 60 days: +0.18 stars. The rating bump came from guests no longer feeling misled.

Listing C — studio, top floor, sloped ceiling. Started at 8 photos, all wide shots. Audit found: photos read as a Craigslist room rental, no lifestyle, no neighbourhood. Changes shipped: kept the existing cover (it was good), added 11 photos — coffee setup, workspace, balcony, two neighbourhood shots, two amenity details, one floor plan. New count: 19 photos. 30-day booking rate: +51%. The bump came almost entirely from getting past the listing's "is this real" threshold.

The pattern across the three: cover swap is the highest-impact change for an existing listing; adding photos to reach 19–22 is the highest-impact change for a short listing; cutting from 30+ to 22 is the highest-impact change for a bloated listing.

One opinionated take

Most listings rank lower than they should because of photo order, not photo quality. The host hires the photographer, gets back 40 beautiful shots, uploads them in the order the photographer delivered them, and never thinks about the sequence again. The photographer ordered for portfolio flow, not booking conversion. The cover lands wherever — sometimes the prettiest shot, sometimes the most representative one, rarely the one that wins clicks against the competing card on the search grid.

Re-order before you re-shoot. It costs nothing and takes an hour. Drop the cover-photo candidate that the photographer picked and run the cover-photo A/B yourself for two weeks per finalist. If you want the listing's booking and click data tracked alongside the photo changes you ship — which cover swap moved the needle, which order change did — that is the kind of thing RentTools is built to surface. And once the photos are working, the next leverage point is the pricing curve underneath them — covered in dynamic pricing for short-term rentals.