Food Photography Review: Chef, Stylist, Marketer

Why food shoots stall in review

Food is the one genre where three different kinds of expertise look at the same pixels and see different problems. The chef sees whether the dish is true: is that the right garnish, is the sauce the right consistency, would a regular recognise it? The stylist sees whether it is composed: height, colour separation, the crumb that got moved. The marketer sees whether it is usable: room for a headline, the brand palette, a crop that works in a square and a banner.

When all three comment on a whole gallery in one email thread, the failures are predictable:

• A chef's "the plate looks wrong" cannot be told apart from a stylist's problem or a retouching problem.
• The marketer's "can we go wider" is a reshoot request that arrives after the set is struck.
• Two people ask for opposite things on the same frame and the photographer becomes the tiebreaker.

There is a cost side too. Every reply that restates a question already asked is unbilled time for the photographer, and every reshoot request that arrives after wrap is either a second day rate or an argument. Food is also perishable in a way other genres are not: the dish, the props and the light are only available together on the shoot day. That is why the order of review, and how quickly the first round returns, matters more here than in almost any other kind of commercial work.

None of this is about taste. It is about who owns which decision, and whether the order they decide in makes sense.

Give each person a lane

Split the review by concern, not by seniority. Each lane has a question, a kind of note that is allowed, and a kind that is not.

The redirect column matters most. A note in the wrong lane is not wrong, but it is routed to the wrong person. If the chef says the sauce looks flat, the stylist decides whether that is a reshoot item or a retouch item. Write that rule into the brief before the shoot, in one sentence per role.

Run the rounds in order: dish, then composition, then use

Sequence beats simultaneity. If all three review at once, the marketer spends time choosing between frames the chef would have vetoed. Three short rounds fix that.

1. Round 1, the chef, dish accuracy. Every frame is a candidate. The chef marks each one keep or drop on accuracy alone. Anything the chef drops is out of the pool, permanently.
2. Round 2, the stylist, composition. Only the survivors go on. The stylist picks the strongest angle per dish and flags fixable problems.
3. Round 3, the marketer, usability. The marketer sees only what is left, typically one to three frames per dish, and picks the hero by placement.

The pool narrows at each step, so nobody reviews more than the previous person left them. In a 60-frame example where the chef cuts a third and the stylist keeps the best one or two per dish, the marketer might see 15 to 20 frames, not 60. That is an illustration of the mechanism, not a measured result.

Make every note about one frame

The most useful habit in food review is refusing gallery-level feedback. "The lighting is moody" is not an instruction. "Frame 14: the shadow under the bowl hides the noodles" is. I made the longer case for this in why per-image comments beat whole-gallery feedback, and it applies doubly when three people are commenting.

Per-frame notes also let you attribute. When a note says "crop tighter" on a frame, you can see it came from the marketer's round, so it is a usability request and not a claim that the dish is wrong. In a shared email thread that provenance is gone by the third reply.

Handle the disagreements you will get anyway

Even with lanes, two reviewers will occasionally want opposite things. Decide the tie-break rule before the first round, not during the argument.

• Accuracy beats aesthetics. If the chef says a garnish is wrong, that overrides a stylist's preference for it. A beautiful frame of the wrong dish cannot go on a menu.
• Composition beats placement. If a crop that suits the banner ruins the plate, the stylist's composition stands and the marketer gets a second frame instead.
• One named decider per project. Usually whoever signs the invoice. Their sign-off closes the round.

The same discipline is what keeps a large fashion team from dragging a selects round into a meeting; the round structure I describe in fashion editorial selects without a group meeting transfers directly to a three-person food review.

What the photographer should set up before the shoot

Most of the work is done before the first frame is shot. A one-paragraph brief, sent with the shot list, should contain:

1. The three (or four) named reviewers and the one question each answers.
2. The order of rounds and the date each closes.
3. The tie-break rule and who the final decider is.
4. How many rounds of retouching are included, and what counts as a reshoot instead.

On the delivery side, one review link per round keeps each person on their own lane. Appruva is built for exactly this shape of work: the client opens a link in the browser, marks up individual frames, makes selects and leaves notes, with no account to create, and a retouched round comes back as a new version of the same frame. Whatever tool you use, the requirement is the same: notes pinned to frames, rounds you can tell apart, no login wall between a chef and the photographs.

Keep the brief short enough that people read it. A single page, sent when the shoot is booked and repeated in the message that carries each review link, does more than a long process document nobody opens. The reviewers should be able to answer "what am I being asked, by when, and what happens next?" from the first two lines of what they receive.

A worked example: a 12-dish menu refresh

Take a restaurant with 12 dishes to photograph, and shoot five candidate angles per dish. That is 12 × 5 = 60 frames, the same pool as above.

The ceiling is what matters. Because the stylist keeps at most two angles per dish, the marketer never faces more than 24 frames, however many the chef let through. Compare that with the three-way email thread, where all 60 frames are open to all three people for the whole time.

The arithmetic also shows where to put your own effort. The chef's round has the most frames and the fewest kinds of decision, so it should be the quickest to send and the strictest about scope. The marketer's round has the fewest frames and the most kinds of decision, so give it the most time on the calendar. Deadlines set in proportion to the decisions, not the frame count, are easier to hold.

Two practical notes from this shape. First, name the reshoot window explicitly: if the chef drops all five angles of a dish, that dish needs a reshoot, and that is only cheap while the set and ingredients still exist. Second, keep a plain record of what each round decided. When a marketer asks three weeks later why a frame is not in the set, the answer should be a note on that frame, not a search through someone's inbox.