Bulho Labs

Goodlight · data · photos

We audited 20,000 holiday-rental photos. 9 in 10 have a flaw, and most of them can be fixed.

We ran 19,964 photos from 681 listings in Portugal through a photo-by-photo audit: what each picture shows, what is wrong with it and what score it deserves. The most common flaw isn't the light. It's leaning walls.

Data from · By Bulho Labs

Goodlight · Infographic

Bulho Labs

90%

of holiday-rental photos have at least one flaw

  • 47%

    have leaning walls or wide-angle stretch

  • 78%

    of listings don't have their best photo as the cover

  • 3%

    of photos got the top score

19,964 photos · 681 listings · Portugal · Nov 2026Download image ↓Social post format ↓
19,964
photos audited, one by one
681
Airbnb and Booking listings
90%
have at least one flaw
3%
scored the top mark (5 out of 5)

01 —

The most common flaw: leaning lines

Almost half the photos (47%) have walls, doors or windows that lean, or the stretched look of a wide-angle lens. Indoors, more than half. That's what happens when the phone is tilted up, or switched to 0.5× to "fit everything in".

Next come framing (beds and sofas cut off, too much ceiling) and colour: yellow or orange bulb light that makes a room look smaller and older than it is. Dark, blurry or blown-out photos are rarer than people think.

Straight from the phone
After straightening

What's wrong with the photos

Share of photos with each kind of flaw. One photo can have several.

  • All photos
  • Indoor photos
  • Leaning lines or wide-angle stretch47%56%
  • Framing (cut-offs, too much ceiling)15%23%
  • Yellow, orange or cold colour16%23%
  • Blown-out windows or highlights7%10%
  • Clutter, visible cables7%8%
  • Dark or gloomy8%7%
  • Blurry, grainy or small9%5%
See the data
FlawAllIndoor
Leaning lines or wide-angle stretch47%56%
Framing (cut-offs, too much ceiling)15%23%
Yellow, orange or cold colour16%23%
Blown-out windows or highlights7%10%
Clutter, visible cables7%8%
Dark or gloomy8%7%
Blurry, grainy or small9%5%

02 —

Room by room

Hallways, living rooms and bathrooms have the most flaws: over 95% of their photos. They're small spaces where the wide-angle lens is most tempting, and 6 in 10 of those photos end up with leaning lines.

Pool and view photos come out best. Outdoors there is plenty of light and no walls to lean.

Photos with at least one flaw, by room

Under each room, the average photo score (1 to 5).

  • Hallway and entranceaverage score 3.2796%
  • Living roomaverage score 3.5896%
  • Bathroomaverage score 3.3495%
  • Bedroomaverage score 3.5095%
  • Kitchenaverage score 3.4693%
  • Dining roomaverage score 3.5790%
  • Gardenaverage score 3.2789%
  • Exterioraverage score 3.4688%
  • Balcony or terraceaverage score 3.5988%
  • Viewaverage score 3.4685%
  • Poolaverage score 3.7580%
See the data
RoomPhotosWith a flawLeaning linesAverage score
Hallway and entrance71796%60%3.27
Living room2,84396%59%3.58
Bathroom1,90595%60%3.34
Bedroom3,70195%53%3.50
Kitchen1,78493%54%3.46
Dining room95690%45%3.57
Garden59489%32%3.27
Exterior1,53788%52%3.46
Balcony or terrace1,53688%47%3.59
View1,59685%23%3.46
Pool43180%33%3.75

03 —

Good photos, rarely great ones

On a 1-to-5 scale, nine in ten photos land on a 3 or a 4: they show the place, but nothing stands out. Only 3% got the top mark and under 7% scored below 3.

Per listing the pattern repeats: in 99% of listings at least half the photos have a flaw, and in 84% at least one photo in four has leaning lines.

Photo scores, 1 to 5

0%
6%
45%
46%
3%
12345

audit score

See the data
ScorePhotosShare
1290.1%
21,2376.2%
38,96144.9%
49,16445.9%
55732.9%

04 —

The gallery is a problem too

In 78% of listings the cover isn't the best photo the listing has. In a third of them the best one sits 11th or later: most people browsing never see it.

Almost eight in ten galleries repeat photos (the same angle two or three times, 2.0 groups on average). And on average 27 of the 29 photos have no caption.

Where the best cover photo is

Position, in the current gallery, of the photo the audit picked as the cover (678 listings).

  • Already the cover22%
  • 2nd to 5th27%
  • 6th to 10th18%
  • 11th or later33%
See the data
PositionListingsShare
Already the cover14922%
2nd to 5th18527%
6th to 10th12118%
11th or later22333%

05 —

Amenities the photos show but the listing doesn't tick

In 93% of audited listings the photos show amenities the host didn't tick in the list: 4 on average. Search filters read that list, so a balcony that only appears in a photo doesn't count when a guest filters for one.

The most commonly forgotten amenities

Share of the 678 listings where the amenity is visible in the photos but not ticked.

  • Books and reading material32%
  • Garden view32%
  • Outdoor furniture31%
  • Dining table29%
  • Private patio or balcony21%
  • Oven17%
  • City skyline view17%
  • Sun loungers16%
  • Sea view16%
  • Room-darkening shades15%
See the data
AmenityListingsShare
Books and reading material21832%
Garden view21432%
Outdoor furniture20731%
Dining table19629%
Private patio or balcony14021%
Oven11617%
City skyline view11417%
Sun loungers11016%
Sea view10916%
Room-darkening shades10415%

06 —

And the rest of the market?

A full audit takes time and money, so we also measured, without a model, the light, sharpness and size of 279,906 photos from 7,787 listings. More than one in five is dark, another one in five is soft, and one in seven was uploaded under 1,200 pixels on the long side.

Automatic measurements on 279,906 photos

No AI: average brightness, fine detail and size of each file.

  • Dark (average brightness under 46%)23%
  • Soft (little fine detail)22%
  • Small (under 1,200 px on the long side)14%
See the data
MeasurementShare of photos
Dark (average brightness under 46%)22.8%
Soft (little fine detail)22.3%
Small (under 1,200 px on the long side)14.3%

07 —

What to fix first

  1. 01 —

    Straighten the verticals

    Shoot with the phone at chest height and upright, not tilted up. If the photo already exists, the perspective tool in Lightroom or Snapseed straightens it in seconds.

  2. 02 —

    Drop the 0.5×

    The ultra-wide lens bends walls and pushes furniture away. Use 1× and step back to the doorway.

  3. 03 —

    One kind of light

    Switch off the yellow bulbs or fix the white balance. Daylight alone gives cleaner colour.

  4. 04 —

    Choose the cover

    The cover is the photo shown in search. Put the brightest, widest room first.

  5. 05 —

    Delete the repeats

    Two photos of the same angle don't show more of the place. They show less variety.

  6. 06 —

    Tick what the photos show

    Go through your photos and tick every amenity you can see: balcony, view, oven, sun loungers.

The studies linking photo quality and cover choice to more bookings are in the Goodlight guide.

Method and limits

Sample: 19,964 photos from 681 listings (Airbnb and Booking) in Portugal, collected in October and November 2026. A vision model (Claude Sonnet) looked at every photo with the same brief: room type, a 1-to-5 score, flaws and contents. We grouped the flaws into seven families; one photo can have several.

The 681 listings are not a random sample: many were picked for the audit because the automatic pre-screen flagged weak photos. 63% of them have photos flagged as weak, against 41% of the other listings in our database. The audit percentages describe these listings, not the whole market. For the market, use the pre-screen numbers.

A vision model makes mistakes too: it can see an amenity that isn't there or count two similar photos as repeats. We don't publish any individual listing, host or photo.

Want to see what we'd find in your listing?

We look through your listing's photos and fix two of them for free. No strings attached.