FoxiPresenting Before-and-Afters Without Overselling
Before-and-after pairs are the fastest way to show someone what image work is worth, and the fastest way to lose their trust. The failure is rarely a lie. It is usually a series of small conveniences: the before is a slightly wider crop, it is displayed smaller, it happens to be the frame where the light was worst, and the caption says nothing about what changed. Each choice is defensible on its own. Together they turn a demonstration into a sales trick, and experienced prospects can smell it. A fair comparison is not weaker than a flattering one. It is more persuasive, because the viewer stops looking for the catch.
Decide what the pair is meant to prove
Most before-and-afters are assembled without anyone deciding what claim they support. Is the claim that you can make a poorly lit phone photo usable? That you can put a product on a clean background consistent with a client's brand? That you can remove distracting clutter without touching the subject? Each of those needs a different pair, and a pair chosen for one claim rarely proves another.
Write the claim as a sentence first, then pick the images. If the claim is about consistency across a set, one pair proves nothing and you need a grid of six finals instead. If the claim is about rescuing bad source material, you need a before that is realistically bad, not theatrically bad.
This also tells you when not to use a comparison at all. Work whose value is judgement rather than transformation, choosing the right frame, the right crop, the right restraint, is badly served by a two-panel image. Show the final and explain the decision.
Match the pair on everything except the change
The comparison is only readable if one variable moves. That means identical crop, identical aspect ratio, identical output size, identical display size on the page or slide, and identical viewing conditions. A before shown at 600 pixels wide next to an after at 1200 will always look worse, and the viewer will read the difference as your work when half of it is scale.
Watch orientation and framing especially. If the after is cropped tighter, some of the improvement is the crop. If that crop was part of your work, say so explicitly rather than letting it pass as retouching. Cropped to a 4:5 portrait for feed placement, background replaced, exposure lifted is a complete and honest description.
Colour management is the quiet one. If the before was exported in one colour profile and the after in another, they will not compare fairly on a browser that handles profiles inconsistently. Export both the same way from the same tool at the same time.
Do not manufacture the before
The most damaging habit in this genre is arranging a worse before. Underexposing on purpose, shooting the original with the room lights off, leaving obvious clutter in a frame you could have tidied, choosing the one out-of-focus frame from a set. It works on a first-time viewer and it fails badly on a second look, which is exactly the look a prospect gives before signing anything.
Use the real starting file. If the real starting file is not very dramatic, that is information. It means your improvement in that case was modest, and you should either present a different case or present that one honestly as a subtle refinement. A modest, clearly true improvement is more credible than a dramatic, faintly suspicious one.
If a prospect ever notices that a before was staged, the conversation stops being about your work. It becomes about what else you might be shading, and there is no good recovery from that inside a pitch.
Label what changed, in plain words
Every pair should carry a caption listing the actual edits. Not adjectives, edits. Background replaced with a neutral studio sweep. Colour balance corrected from mixed tungsten and daylight. Power cable and floor scuff removed. Crop changed from 3:2 to 1:1. Nothing else altered. A caption like that takes thirty seconds to write and it answers the question the viewer was already forming.
There is a category of change that should never appear in a caption because it should never happen. Editing changes the photograph, not the subject. Lighting, colour, background, framing and distracting clutter are fair game. A product's real colour, shape, size, material or condition is not. Neither is a person's actual appearance, a room's real layout and fittings, or the state of anything a buyer will physically receive. If a pair implies the subject itself improved, the pair is misleading no matter how it is captioned.
This constraint is also a selling point when you explain it to a client. Tools like Foxi AI work on a photo you already have and are built around preserving the real subject while changing background, lighting and presentation. Being able to say plainly which parts you will and will not touch is a reassurance to any client whose category is regulated or whose customers return things.
Be careful about outcome claims
The temptation next to a good pair is to attach a result. Sales up, engagement up, bookings up. Unless you have clean measurement isolating the image change from price, seasonality, spend, placement and everything else that moved in the same week, that number is not evidence and attaching it to your portfolio makes the whole page less trustworthy.
Nobody can promise that a better photograph produces more revenue. What you can honestly say is what the work did to the asset itself: it is consistent with the brand's other images, it is legible at feed size on a phone, it meets the placement's requirements, it no longer shows a distracting background. Those are verifiable and they are enough.
If a client volunteers a result, quote them as a client statement rather than presenting it as your measured outcome. The attribution difference is small on the page and large in a dispute.
Formats that survive a slide, a page and a phone
On a slide, side-by-side beats a slider every time, because a projector plus a live audience gives you no chance to interact. Put both images at the same height with a thin gap, label them with small type in the same position on each, and leave the caption on the slide rather than saying it aloud, since decks get forwarded without you.
On a web page, a slider is fine if the handle starts at the centre and the two images are pixel-identical in dimensions. Sliders that start fully on the after, or that autoplay, feel like an advertisement. Provide a static side-by-side fallback for anyone who does not drag.
On a phone, side-by-side halves the width and both images become too small to judge. Stack them vertically with clear labels instead, or show the after full width and the before as a smaller inset with a caption. Whatever you choose, test it at real phone width before publishing, because a comparison nobody can see clearly is not a comparison.
- Same crop, same aspect ratio, same export settings, same display size
- Use the genuine original file, never a deliberately worse one
- Caption the specific edits, not adjectives
- Never imply the subject itself changed, only the photograph
- Stack vertically on phone; side-by-side on slides; slider only on web
Frequently asked
Is it acceptable to tidy a set before shooting the before image?
Yes, and you should. The before is meant to represent the realistic starting point, not a worst case you engineered. Shoot the scene as you normally would, then use that frame as the before. What is not acceptable is degrading the original to make the edit look larger.
How many pairs should a case study include?
Usually one or two, plus a grid of finals. Long strings of pairs stop being evidence and start reading as filler, and each additional pair is another chance for a viewer to spot an inconsistency. If the point is consistency across a set, skip pairs entirely and show the set.
Can I show a before-and-after of a client's product photo publicly?
Only with permission. The before is often the client's own file and may be commercially sensitive, and some clients specifically do not want customers seeing that their listing images were edited. Ask before publishing, and keep the request in writing.
What if the client's original photo is genuinely unusable?
Say so at intake rather than proving it later with a dramatic pair. Editing improves lighting, colour, background and framing; it cannot recover detail that was never captured, and it cannot invent a scene that was not photographed. Telling a client early that the input will not produce the output they described is a better conversation than showing them a heroic rescue afterwards.
Try it on your own photo
Foxi runs this kind of edit in about a minute — upload a photo, pick a look or describe the change you want, and see the result before you pay for anything.