A product tag is the best possible answer to "where is that from", for exactly one item. It was added by the person who owns the garment or sells it, it points at a specific product page, and it is correct.
Then you look at the rest of the outfit and there is nothing.
What does a tag actually give you?
The exact product, which no visual search can match.
A tag names the item. Not a description of it, not something that looks like it, but the actual product page with the actual name, the actual composition and the actual sizes. If you want that specific garment, the tag has answered your question completely.
Snagfit works differently by design. It reads the garments in a photo, describes what each one looks like, and searches shopping listings for that description. What it returns is where a garment is sold or the closest thing you can actually buy, which is a weaker claim than a tag makes and an honest one, as why Snagfit shows close matches sets out.
So on the one tagged piece, the tag wins. The question is everything else in the frame.
Why is most of an outfit untagged?
Three reasons, and the first is the main one.
Tagging follows money. A creator tags what they are paid to feature or earn commission on. The pieces they bought themselves three years ago have no commercial reason to be tagged, so they are not. That is why the shoes are linked and the jumper never is.
Tagging takes effort per item. Each tag has to be placed and each product found in the platform's catalogue. On a post with seven garments that is a real amount of work for no return on six of them.
Platforms cap it. Most limit how many products can be tagged in a single post, so even a creator who wanted to tag everything cannot always do so.
The practical shape of this: a typical outfit post has one or two tagged items out of six or seven visible garments. The tagged ones are usually the most expensive or the most recently acquired, which are not necessarily the ones you wanted.
What happens when a post gets reposted?
The tags stay behind, and this is the failure people hit most often.
Product tags are data attached to the original post on the original platform. They are not part of the image. So a screenshot, a download, a repost, a save to a different app, or a copy shared in a message all carry the picture with none of the tag information.
Which means that if you saw the outfit anywhere other than the source post, the tags may well exist and you cannot see them. There is no indication in the image that they are there.
Sometimes you can find the original. A search on the creator's name and the date, or scrolling their profile back to the post, gets you to the tagged version. Often you cannot, particularly when the repost stripped the attribution too.
That is the situation where searching the image is the only route left, and it is a common one. Most saved outfit photos in a camera roll are reposts of something.
Do tags stay accurate over time?
No, and nobody maintains them.
A tag points at a product page and keeps pointing there regardless of what happens to the product. Eighteen months later that page may show the item as sold out, discontinued, replaced by a new season's version, or removed entirely.
The tag has not broken. It is doing exactly what it always did, which is record what the person was wearing at the time. It was never a live shopping link and it was never going to be updated.
This is worth knowing because a dead tag still tells you something valuable: the product name. With that, a resale search or a search across other retailers becomes straightforward, which is the whole approach in finding a piece that has sold out.
Does a tag point at the best price?
It points at one retailer, and that retailer is usually the one the creator has a relationship with.
Many product tags are affiliate links, meaning the creator earns a commission on purchases made through them. That is a normal and disclosed arrangement, and it means the link goes to whichever retailer runs the affiliate programme rather than to whichever is cheapest.
For an item stocked by several shops, that can be a real price difference. The same garment at a brand's own site, at a department store and at a marketplace seller can carry three prices, for the reasons in why the same item has different prices.
So a tag is worth following for the product name and worth checking against other retailers before buying. A search across retailers does that comparison as a side effect, which is one of the practical arguments for running both.
Snagfit's own buy links work the same way and it is disclosed on the legal page. Where an affiliate relationship applies, the link is wrapped so a purchase can be attributed. It does not change the price you pay.
Are brand posts better tagged?
Usually, and they are the case where tags come closest to covering a full outfit.
Every garment in a brand's own photograph is their own product. There is no commercial reason to leave any of it out, and the catalogue is right there, so brand posts often tag most of what is visible.
The limitation is that a brand only tags its own pieces. Anything styled in from elsewhere, which on a well-styled shoot is often the shoes, the bag and the jewellery, stays untagged. So even a fully tagged brand post leaves the accessories open, and accessories are what people most often want.
Retailer posts behave similarly and sometimes better, since a multi-brand retailer stocks more of what is in the frame.
What about the link-in-bio page?
The same trade-off as a tag, spread across a whole account.
Many creators maintain a page of links to things they have worn, organised by post or by date. It is more durable than a tag, since it survives reposting and it is findable from the profile rather than from a single post.
It has the same coverage gap for the same reason. The page is built around sponsored and affiliate items, because that is what pays for maintaining it. So a post with seven garments has one or two on the page.
It also has a staleness problem of its own. These pages accumulate over months and the links are rarely pruned, so a page can carry dozens of entries pointing at products that no longer exist.
Worth checking before you search, since it costs thirty seconds and occasionally answers the whole question. Not worth relying on.
Does any platform tag automatically?
Some surfaces attach shopping information without the creator doing it, and it is worth knowing which you are looking at.
Platform-generated shopping links, where a system identifies a product in an image and attaches a link, exist on several surfaces. These are visual matches rather than declarations by the owner of the garment, so they carry the same uncertainty as any visual search: they may be the item, and they may be something that looks like it.
That is a different thing from a creator-added tag, and the distinction matters when you are deciding how much to trust it. A creator-added tag is a statement by somebody who owns the garment. A platform-generated one is a guess by a system, presented in the same visual style.
Where you cannot tell which you are looking at, treat it as a strong suggestion and check the product page against the photo before buying, which is what reading a shopping result is for.
What does using both look like?
One tap and one upload, and it takes under a minute.
Follow the tag for whatever is tagged. That gets you the exact product with no ambiguity, and it is the strongest answer available.
Screenshot the frame and search it for everything else. Snagfit returns up to eight garments per photo, so a single upload covers the rest of the outfit. The result on each is a description and somewhere it is sold, which is weaker than a tag and much better than nothing.
Then compare. If the tagged item also appears in the search results at a different retailer, you have a price comparison you did not have to do by hand.
The one thing worth deciding first is what you actually want, because building an outfit around one piece makes the case that buying all seven garments from one photo is how wardrobes fill up with clothes nobody wears. The tag tells you about one piece. The search tells you about the rest. Neither tells you which of them you needed.