Troubleshooting Guide

Food Delivery Photo Rejected? Here's How to Fix It

Every common rejection reason on DoorDash, Uber Eats, Grubhub, Deliveroo, Rappi, iFood and Glovo — and the specific fix for each. Resubmit approved in under 5 minutes.

Fix photos with AI Score my photo free

Most common reason

Wrong resolution

Fastest fix

Re-export

Universal size

1600 × 1600 px

Approval time

24 – 72 hours

Rejected menu photos cost restaurants real money: every day a dish sits without an approved photo, it converts lower in the delivery app feed. The good news is that 95 percent of rejections come from a short list of technical mistakes that can be fixed by re-exporting the file — no reshoot required.

This guide covers every rejection reason used by DoorDash, Uber Eats, Grubhub, Deliveroo, Rappi, iFood and Glovo, grouped by how to fix them.

Technical rejections (fix in 2 minutes)

These are automated checks that reject the file at upload before a human sees it. Fixing them requires re-exporting the image, not reshooting.

Resolution too low: export at 1600 × 1600 px to clear every platform minimum
Wrong aspect ratio: crop to exactly 1:1 (square) — most platforms require this strictly
File too large: save as JPEG at 85 percent quality to get under 5 MB
Wrong format: convert HEIC, WebP or AVIF to JPEG or PNG
Wrong color space: convert Adobe RGB or P3 to sRGB before export

Content rejections (fix in 5 minutes)

These rejections come from manual review. A human looked at the photo and flagged it for a content issue that violates the platform's authenticity or composition policies.

Price overlay on photo: remove the price text entirely — platforms show prices separately
Watermark: crop out the watermark or use a non-watermarked source
Hands, phones, or cutlery being held: reshoot or crop out the offending elements
Multiple dishes in one photo: crop to show only the dish being listed
Branded napkins or tablecloths: swap for neutral background (white, wood, slate)

Authenticity rejections (harder to fix)

Platforms actively reject stock photography, AI-generated photos that look unrealistic, and photos that do not match the dish description. These rejections can trigger account-level flags if they repeat.

Stock photo detected: replace with an actual photo of the dish your restaurant serves
Generic AI photo that does not match the menu item: generate from an actual photo of your dish
Photo does not match description: make sure ingredients shown match what the customer receives
Different portion size than reality: reshoot or generate from a realistic input

Quality rejections (requires reshoot or AI enhancement)

Some rejections cannot be fixed with a simple re-export — the underlying photo has issues that need enhancement.

Blurry: use a tripod, enable stabilization, or reshoot; AI can sharpen slightly but not fully recover motion blur
Underexposed (too dark): reshoot in better light, or use AI relighting
Overexposed (blown highlights): reshoot with less direct light; AI can recover mid-tones but not clipped highlights
Green or yellow tint from kitchen fluorescents: reshoot in daylight, or use AI white balance correction
Oversaturated HDR look: re-edit with natural saturation — platforms flag HDR-style photos

Platform-specific rejection patterns

Each delivery app has idiosyncrasies that cause rejections unique to them.

DoorDash: most strict on 1:1 aspect ratio and 1000 × 1000 px minimum
Uber Eats: stricter on authenticity — will flag photos that look too perfect or overly styled
Grubhub: rejects dark moody photography; prefers bright, appetizing lighting
Deliveroo: stricter on oversaturation and HDR effects
Rappi: rejects anything below 1000 × 1000 px at upload; manual review strict on stock photography
iFood: highest minimum at 1280 × 1280 px; photos sized for Rappi will be rejected
Glovo: strict on branded tableware and competing dishes in the frame

The universal export that prevents 90 percent of rejections

One export setting clears almost every automated rejection reason across all major delivery apps.

Resolution: 1600 × 1600 pixels
Aspect ratio: 1:1 (square)
Format: JPEG at 85 percent quality
Color space: sRGB
File size: typically 600 KB to 1.5 MB (well under 5 MB limit)
Works on: DoorDash, Uber Eats, Grubhub, Deliveroo, Just Eat, Rappi, iFood, Glovo, Swiggy, Zomato, Wolt, foodpanda

Frequently asked questions

Why did DoorDash reject my photo?

DoorDash rejects photos for resolution below 1000 × 1000 px, non-square aspect ratio, file over 5 MB, watermarks, price overlays, stock photography, blurry or dark images, or photos that do not match the item description. Most rejections fix with a re-export at 1600 × 1600 px square JPEG.

Can I appeal a rejected delivery photo?

Yes on every major platform. DoorDash, Uber Eats, Grubhub, Deliveroo, Rappi and iFood all allow resubmission after changes. You do not need to formally appeal — just upload a corrected version and it goes back through review.

How many times can I resubmit a rejected photo?

There is no hard limit, but repeated rejections on the same item can flag the account for partner manager attention. Fix the specific rejection reason before resubmitting — submitting the same file twice will get rejected twice.

Do photo rejections affect my restaurant rating?

Rejections do not directly affect customer-facing ratings, but missing or pending photos hurt conversion because the item shows a placeholder or generic image. Indirectly, this affects sales and may reduce algorithmic placement.

How long do I have to fix a rejected photo?

There is no deadline, but the item stays live with a placeholder image until you upload an approved photo. Conversion on items without approved photos typically runs 30 to 60 percent below items with approved photos.

Are AI-generated food photos rejected?

AI-generated photos that are realistic and match the actual dish are generally accepted. Unrealistic AI photos (exaggerated portions, impossible plating, stock-looking aesthetic) get rejected under authenticity policies. The safest approach is AI enhancement of a real photo of your dish, not fully synthetic generation.