Ten mistakes people make filming their own product

A product video is an asset with a working life. It either keeps selling the object for a year or it dies the week the promotion ends, and which of those happens is settled on the day you film rather than in the edit.

Ten is reachable on a subject this narrow. Filming something you sell is a repeatable job, and repeatable jobs fail in specific ways rather than vague ones. Every item below has a cost attached, and the cost is usually a reshoot nobody budgeted for.

Only one of the ten involves buying anything, and the thing you buy is a piece of card.

One thing this list is not: the marketplace's own image rules. Those are hard rules rather than creative choices, they differ by marketplace and category and they change, so they get checked against the live specification for your exact destination before a shoot rather than remembered. Cut-out images for marketplaces is that side of it. Everything below is the creative and production half, where the judgement is genuinely yours.

Ten ways a product shoot loses money

1. Moving the light between products

A range photographed across a week, with the lamp nudged for the awkward item and again for the shiny one, comes out looking like a range from four different shops. The bill is not one poor photograph, it is that correcting a single item means matching a lamp position nobody wrote down, so you reshoot the set instead. Mark where the lamp stands on the floor and where the camera stands on the tripod, and shoot the whole range in one sitting so the set matches. For an ordinary item that holds without exception. A genuinely reflective surface, chrome, glass or high gloss, is the one case where the fixed position fails it, throwing a hotspot or a black edge no grade will rescue; there you move the light for that item alone and match the result back, the same exposure, background and look, so what stays constant is the picture the range produces rather than the lamp's coordinate.

2. No consistent position or scale

Each shot is fine on its own and the grid is chaos, because the object sits high in one frame and small in the next. A customer reads that inconsistency before they read any single image, and they also cannot compare two of your own products, so they pick whichever photographed best rather than whichever suits them. Decide one framing, mark on the surface where the object goes, and shoot everything to that mark.

3. Filming the product without ever handling it

The object sits still on a surface, the camera moves around it, and nothing in the footage tells anybody what it weighs, how the finish feels or what the lid does. Those are the signals a buyer is actually reading, and none of them survive a shot of a stationary object. Get hands into the frame lifting it, opening it and setting it down; "What changes when an agency holds your product" is the same gap from the other side, which is that whoever makes your video should be asking you to send them one.

4. Shooting one aspect ratio for something that needs two

A single wide master cropped to upright either loses the product off the sides or strands it in the middle of the frame with bars above and below. That hands whoever runs your advertising one lever instead of two, and the upright placements are frequently the cheaper ones to buy. Film the key shots twice, once wide and once upright, while everything is still set up; "Feed and Story are two assets" explains why the two are not versions of each other.

5. Burning the offer into the picture

The moment the promotion changes, a perfectly good video is unusable, and the object in it has not changed at all. You have tied the working life of the asset to the working life of the offer, which is the shortest-lived thing in the business. Keep the offer in the caption, on the landing page, or in an end card that swaps without touching anything before it; "On-screen text that answers a question" covers what belongs burned in and what does not.

6. No text-free master

This is the same fault one step earlier and it costs more. Even sensible on-screen text, the product name, the feature callouts, the sizing, means a second language, a second market or a second season is a rebuild rather than an export. Ask for a graded version carrying no text at all and keep it with the project file; "Getting your project files back" is about agreeing that before the work starts rather than after.

7. Room noise on an unboxing

The crinkle of the paper, the resistance of the seal and the click of the lid are most of what tells somebody the thing is well made, and all of them are quiet sounds. A fridge, a road and a room behind them means the whole recording has to be cleaned, and removing quiet detail is precisely what cleaning does. Film the unboxing at the quietest hour you have, with the microphone close and everything else switched off; "Noise removal, and when to stop" sets out the trade.

8. Rushing the moment the box opens

The pause before the lid moves is the event. Filmed quickly it becomes a transition, so you have cut past the only part anybody came for, on the piece of content a nervous buyer studies hardest. Slow the hands down, film the opening more than once, and keep all of it including the resistance; "Unboxing videos that are not an afterthought" argues for filming it first rather than last.

9. One hook

With a single opening there is nothing to test against, so when the advert underperforms you cannot tell whether the product, the offer or the first two seconds is the fault. Film the opening several ways in one sitting, while the light and the object are still where you left them, and let the platform tell you which one works. Filming several hook takes and sending them with the footage is this studio's habit, so the winner is chosen on performance rather than on whose favourite it was; "Film three hooks, not one" is the method.

10. No reference card

Photograph a colour reference card once at the start of a set and every image afterwards can be corrected against the same neutral. Skip it and colour drifts across the set, so the jumper on the listing is not the jumper that arrives, and somebody opening a parcel containing a shade they did not order sends it back. A return is the most expensive unhappy customer available, and this is the one item on the list that involves buying anything; "Matching colour across cameras" covers the same correction across a shoot with two cameras on it.

Which two do you fix first?

The lamp position and the reference card, because both apply to every item you will ever photograph rather than to one video.

Everything else here is per shoot. Those two are per business: set them once, write down where the lamp goes, and every product you add afterwards joins a set that already matches. If you photograph a range at all, that is the best return available on an afternoon.

What is not worth doing?

Building a set. It is the thing people reach for first, and for most catalogues it is a way of spending a weekend producing something that will not match the next item you add to the shop.

Also not worth it: filming a video for a line you sell rarely. Photograph it properly, put it on the listing, and spend the filming on the product that pays the rent. A video earns its keep when the same footage cuts into an advert, a listing clip and a social post, and that arithmetic only holds for products with volume behind them.

Work out which products those are before anything else here applies to you. Your own sales figures already know, and the answer decides how much of this list you ever need to act on.