We are publishing this page with an empty hand and saying so. Both runs we hold submitted a placeholder address rather than a Groupon URL, so we have no evidence about this service either way - and "untested" is a different claim from "blocked", which is why this page keeps them apart.
Both rows we hold submitted the same placeholder address - the fictional example URL that appears across this whole family of exports - and both returned http 404, which is what a fictional address returns. That tells you the exporter works. It tells you nothing whatsoever about Groupon, and this page does not pretend otherwise.
Two other pages in this catalogue report real refusals - attempts that were made and turned away. This one reports no attempt at all. Those are different facts and they lead to different decisions: a blocked source needs a workaround, an untested one needs a test. We keep the words apart because collapsing them would quietly overstate what we know.
On 12 August 2026 we fetched Groupon's robots.txt and a search page ourselves, with a browser user-agent. Both returned HTTP 200 - the search page came back at roughly 890 KB carrying deal links and the search term throughout. So from where we sit the site is reachable, which does not match the note in our own service catalogue. We report both rather than choosing the tidier one.
This is the finding that should shape your job. Groupon's robots.txt carries no blanket exclusion for the general group - it opens with allow everything and then lists 118 specific rules. One of those rules is Disallow: /search*. Checked with a proper matcher rather than a prefix test: a search URL is disallowed for the general group, while a /deals/ path is allowed. Aim at deal pages.
The same file gives a blanket exclusion to nine named crawlers - the well-known AI-training and archival agents - and carries a machine-readable content signal declaring that search indexing is welcome and AI training is not. None of that applies to a person collecting deal data for their own analysis, but it is the site's stated preference and worth reading before you design around it.
Billing follows rows actually returned, so a run that comes back empty is not a charge. That makes a small job the sensible next step - and makes it unnecessary for us to guess on your behalf. Submit a few deal URLs, look at status, and you will know more about this service in five minutes than this page can tell you.
Read from the header row of a real Groupon export, in sheet order. Every description below is about what the column is for. Where our other catalogue pages quote fill counts from a run, this one cannot: no run of ours ever reached the site.
status were the only fields carrying anything - and what they carried was the placeholder address we sent.sku_code.http 404, on both rows, from a placeholder address that does not exist. Reconcile on this column - and on this service, treat it as the first thing to read.Seventeen columns per row · CSV, JSON or Excel
The honest summary of this table is that it describes a shape, not a result. The seventeen columns are real - they come from the header row of a genuine export and are byte-identical to the schema our Uline, CDW, Waxie, Otto, Newegg, Decathlon, Menards and NapaOnline services use, so an importer written against any of those will accept a Groupon file unchanged. But no run of ours ever addressed groupon.com: both rows we hold are 404s against a fictional placeholder address. We therefore cannot tell you which columns Groupon populates, how a deal price is represented, or whether a run succeeds at all. Several sibling pages on this site do carry full price data on this exact schema, and we have deliberately used none of their figures here - a deal marketplace is a different catalogue from a packaging distributor, and borrowing a number would be inventing evidence.
This is the one page in the catalogue where the obvious next step is yours rather than ours. Submit a few /deals/ pages, read status, and you will know what we do not. Billing follows rows returned, so an empty run costs nothing.
The site's robots.txt asks the general group not to crawl /search, while deal paths are allowed. That happens to be the opposite of the example input recorded in our own service catalogue, so if you were about to copy that example, this is the paragraph to read twice.
If what you need is retail pricing rather than Groupon specifically, our Best Buy, CDW and Uline services returned real money columns from their own runs, and each of those pages states exactly what it measured.
The column list is stable across this whole family of services. Write and test your loader today against a sibling export and point it at Groupon later without a rewrite - the header row is the same seventeen names in the same order.
One time, on signup. No card, and nothing to cancel afterwards.
Billed on rows actually returned, which is what makes testing this service cheap: if nothing comes back, nothing is charged.
Credits do not expire on a monthly cycle, so nothing drains while you decide whether this source is worth a trial.
LIVESCRAPER10 at checkout.One category URL returned 99 priced rows in our run, on the identical schema, with nothing blocked - the clearest working example in this family.
A consumer catalogue where one collection URL fanned out into 66 rows, each a size and colour with its own stock status.
The other page in this catalogue whose runs never left the placeholder - worth reading beside this one, since it draws the same distinction.
The full catalogue, including every other retailer and marketplace we cover on this schema.
Nothing has been collected here, so the question is theoretical for us. It is not theoretical for you, and the site has published a stated preference worth reading.
Prices, offer titles and merchant names on a publicly reachable listing are ordinary commercial facts, and reading them at a considerate pace is what a shopper does by hand. Nothing in the seventeen columns describes a person: no name, no email address, no phone number, no account. Deal listings do name local businesses, which are companies rather than individuals - but if you plan to republish merchant details, that is a licensing question rather than a scraping one, and it is yours to answer.
What is specific to this site is its robots.txt, which we could read: it answered a plain request with HTTP 200 on 12 August 2026. Read with a proper group matcher, the general group carries no blanket exclusion - it allows everything and then lists 118 specific rules - but one of those rules asks crawlers not to fetch /search paths, while deal paths are allowed. The example input recorded in our own service catalogue for this scraper is a search URL, so we are flagging the mismatch rather than reproducing the example as a recommendation. The same file also gives a blanket exclusion to nine named AI-training and archival crawlers and carries a machine-readable signal welcoming search indexing while declining AI training. We report those as the site's stated preferences.
Our own terms are the same as on every other service here. Publicly available pages only, nothing behind a login, no third-party trackers on the data layer, and exports auto-delete after 30 days. Billing follows rows returned, so a run that comes back empty is not a charge.
http 404 - which is what a fictional address returns. We have no evidence about this service in either direction./search paths, and we verified that with a proper matcher: a search URL is disallowed for the general group while a /deals/ path is allowed. Note that the example input recorded in our own catalogue is a search URL, which is exactly why we are pointing this out.price holds before computing a saving from it. That is a caution about reading the data, not a claim about it.Two runs, neither against Groupon, and a direct check that found the site answering normally. Submit a few deal URLs, read the status column, and you will know more than this page can tell you - and an empty run bills nothing.
Groupon is a marketplace for local deals and discounted offers, and this page is published in an unusual state: we have no evidence about the service at all. Our export folder holds two runs and two rows, and both of them submitted the same fictional placeholder address that appears across this whole family of exports rather than a Groupon URL. Both returned a 404, which is what a fictional address returns. So the runs demonstrate that the exporter works and demonstrate nothing whatsoever about Groupon. We could have described the seventeen columns in the confident voice used on our other catalogue pages and let the reader assume; instead this page opens by saying that no request of ours has ever reached the site.
That distinction is the reason the page exists. Elsewhere in this catalogue there are two services whose runs were genuinely refused: attempts made at a real URL, turned away, with timestamps and a status value recording it. Those pages recommend a workaround or a substitute. This one cannot, because an untested source and a blocked source are different facts that lead to different decisions. Collapsing them would be the easy move and the dishonest one. What we can add is a direct check: on 12 August 2026 we fetched Groupon's robots.txt and a search page ourselves, with a browser user-agent, and both returned HTTP 200 - the search page at roughly 890 kilobytes, carrying deal links and the search term throughout. Our own internal service catalogue, meanwhile, marks this scraper as needing a residential proxy. Three sources, not in agreement, all reported here.
The robots.txt is worth reading before you plan a job, because it contains a specific instruction that cuts against the obvious starting point. Parsed with a proper group matcher rather than a prefix test, the general group carries no blanket exclusion at all: it allows everything and then lists a hundred and eighteen specific rules. One of those rules asks crawlers not to fetch search paths, and we confirmed the outcome both ways - a search URL is disallowed for the general group, while a deal path is allowed. The example input recorded in our own service catalogue for this scraper happens to be a search URL, so anyone copying that example would be starting on the one path the site asks the general group to avoid. Point the job at deal pages instead. The same file also gives a blanket exclusion to nine named AI-training and archival crawlers, and carries a machine-readable content signal declaring that search indexing is welcome and AI training is not; those are the site's stated preferences and we pass them on rather than interpreting them for you.
So the practical advice on this page is unusually short. Test it. Billing follows rows actually returned, which means a run that comes back empty is not a charge, and a handful of deal URLs will tell you in five minutes what we cannot tell you at all. If the answer turns out to be that it works, you have a source on the same seventeen-column schema as everything else here and an importer that already fits. If it does not, you have lost nothing. And if what you actually need is retail pricing rather than Groupon specifically, our Best Buy, CDW and Uline pages each report real fill counts from real runs. Publicly available pages only, no third-party trackers on the data layer, exports auto-delete after 30 days, and your first 500 rows are free with no credit card.