The legal research market has spent the last decade convincing lawyers that buying case documents online is a routine transaction, like ordering office supplies. A person types a citation into a search bar, clicks pay, and a PDF arrives with the quiet efficiency of a vending machine dispensing a soda. The comparison flatters the experience. A vending machine cannot sell a defective can. The same cannot be said for every purveyor of legal filings, where the gap between what the thumbnail promises and what the download contains can swallow a filing deadline whole.
Purchasing a case document, a brief, a complaint, or a full docket sheet from a commercial vendor is a gamble wrapped in a convenience fee. Some vendors have cleaned up their acts, digitizing records with genuine care. Others run on databases that were last audited during the Bush administration, offering scans so dark the text reads like a code, or worse, offering documents that are not actually the case a person asked for at all. A lawyer who skips the verification step is not saving time. That lawyer is borrowing time from a future disaster.
The core problem is that the buyer cannot inspect the goods before purchase. The interface shows a title, maybe a case number, sometimes a preview image that is too low-resolution to read. The actual quality of the document remains a mystery until the money has left the account. This makes the pre-purchase checklist not a formality but the entire ballgame. Nine features deserve scrutiny before any payment screen gets the green light.
The vendor's source is the whole ballgame
The first question is not about the document itself but about where it lives before the sale. A vendor that pulls records directly from the PACER system or from an official court clerk's office is working with primary material. A vendor that pulls from another commercial database, which itself pulled from a third database, is working with a copy of a copy. Each generation of copying introduces the possibility of error, whether it is a dropped page, a mislabeled attachment, or a scan that loses the margin where a judge's handwritten note appeared.
Reputable vendors state their sourcing plainly. They say "court records" or "PACER" and they mean it. The murkier operations use phrases like "comprehensive database" or "verified legal content," which translates to a server farm full of files that nobody has checked in years. A person should look for the sourcing disclosure before looking at the price. If the source is not stated, the assumption should be that the source is a landfill.
There is also a difference between a document that was scanned by the court and a document that was scanned by the vendor. Court scanners have standards. Vendor scanners often work with whatever machine is closest to the coffee pot. The resulting file may be missing the court's internal pagination or may have been run through optical character recognition software that garbled every third word. The vendor's scan is not the official record. It is an interpretation of the record, and interpretations can lie.
Page count sanity versus the official docket
Every case document has a page count that exists in the official record. A person can usually find this number on the court's own public access terminal or in the docket text that accompanies the filing. Before buying, a person should know roughly how many pages the document should contain. A 30-page appellate brief that a vendor lists as 12 pages is not a bargain. It is a truncation.
Vendors that charge by the page create a perverse incentive to list page counts that match their own internal file, not the actual document. If the vendor's file is missing the final five pages of a brief, the vendor's system will happily report the document as being 25 pages long, because that is what the vendor has. The buyer only discovers the deficit after purchase, usually while reading a sentence that cuts off in mid-argument and then jumps to a page from a different case entirely.
Checking the page count against the official docket takes about ninety seconds. That ninety seconds is the cheapest insurance available in the entire legal research industry. A person who skips it is making a statement about how much their own time is worth, and also about how much they enjoy redoing work.
The caption on the screen must match the caption on the file
Case captions are like snowflakes in that no two are exactly alike, and like snowflakes in that they are easy to confuse when viewed from a distance. A search for one case often surfaces a different case with a similar name, or the same case in a different procedural posture. The complaint in a matter is not the same as the motion to dismiss in that matter, which is not the same as the order granting that motion. A vendor's search algorithm does not always appreciate these distinctions.
The verification step here is to read the full caption, including the case number, the court name, and the filing date, before paying. A person should read it out loud, if necessary. The case number is the most reliable identifier. If the vendor's listing shows a case number that does not match the docket number a person has been tracking, the document is wrong, regardless of how similar the party names appear.
Duplicate filings complicate this further. Many cases have multiple documents with nearly identical captions, such as an amended complaint and a second amended complaint. The difference between them can be a single added paragraph, but that paragraph may be the entire reason a person needs the document. Buying the wrong version is not a near miss. It is a total failure dressed up as a file download.
Docket text reveals the procedural moment
Every filing on a docket has a text entry that describes what the filing is. This docket text is the metadata that gives the document its meaning. A vendor listing that says "MOTION for Summary Judgment" might be accurate in the narrowest sense, but the docket text may reveal that the motion was denied, withdrawn, or stricken. The document itself is a snapshot of an argument at a specific moment. The docket text tells a person what happened after that moment.
A responsible buyer checks the docket text before purchasing the underlying document. This is not always possible on the vendor's site, which may only display a truncated version of the text. In that case, the buyer should check the court's own public access system first, note the exact docket entry number, and then match that entry number to the vendor's listing. If the vendor's listing does not show an entry number, the risk of buying the wrong filing goes up considerably.
This matters because the timing of a filing is often its most important attribute. A brief filed before a key Supreme Court decision is a historical artifact. The same brief filed after that decision is something else entirely. The docket text, with its date stamp, tells a person which artifact they are holding.
Image quality determines whether the file is usable
A document that cannot be read is not a document. It is a digital paperweight. Vendors differ wildly in the resolution of their scans. The good ones produce files where every word is crisp, where footnotes are legible without magnification, and where handwritten notations appear as clearly as they did on the original. The bad ones produce files that look like they were faxed twice, then photocopied, then scanned on a machine from 1998.
Vendors rarely disclose the resolution of their scans on the listing page. A person often has to rely on the preview function, which may show only the first page. A first page that looks clean does not guarantee that page 40 is legible. But a first page that looks grainy or dark is an immediate red flag. The rest of the document will almost certainly be worse.
The format matters too. A searchable PDF is worth its weight in gold because a person can copy text, search for terms, and highlight passages. A scanned image PDF offers none of that. The vendor will usually indicate whether the file is searchable, but the indication can be buried in the product details. A person who needs to quote from the document should insist on searchable text. A person who only needs to read it once can tolerate an image file, as long as the image is sharp.
Delivery speed is a feature until it is a lie
Most vendors advertise instant delivery, which is true in the sense that the download link appears immediately after payment. The deception is that the instant download is only instant for documents the vendor already has in stock. Documents that require a special request, such as older filings, sealed records that have been unsealed, or exhibits that were never digitized, may take days. The vendor's site often hides this distinction until after the purchase.
A person with a deadline should check the delivery estimate before paying. If the site says "available for immediate download," the document should be in the vendor's possession. If the site says "delivery within 2-3 business days," the vendor is going to fetch it from somewhere else, and the timeline is a guess, not a promise. The workaround is to look for the document on the court's own system first. If it is there, the vendor's delivery estimate is irrelevant.
There is also the question of what "delivery" means. Some vendors email a link. Some provide a download page that expires after 24 hours. Some send a file that requires proprietary software to open. A person who needs to share the document with a colleague or opposing counsel should verify that the file format is standard. A proprietary format is a trap disguised as a feature.
Pricing transparency means no surprise add-ons at checkout
The listed price for a case document often bears little resemblance to the final price at checkout. Vendors add fees for access, for delivery, for "certification," and for the privilege of using their platform. These fees appear in small type or not at all until the payment screen. The practice is legal, but it is also a signal. A vendor that hides its pricing structure is a vendor that treats its customers as marks.
The comparison should be against the cost of getting the document directly from the court. PACER charges ten cents per page, capped at three dollars per document. Many state courts charge nothing at all for public records. A vendor that charges twenty dollars for a document that costs three dollars from the source is charging for convenience, which is legitimate. A vendor that charges twenty dollars and then adds a ten dollar "processing fee" is charging for something else entirely.
The buyer should total the cost across the entire transaction, not just the line item. A document that costs eight dollars with no fees is cheaper than a document that costs five dollars with a four dollar service charge. The arithmetic is simple, but the presentation is designed to obscure it.
Return policies exist for a reason, and the reason is bad documents
A vendor that offers no refunds for defective documents is telling the buyer, in effect, that defective documents are common enough to require a policy. The vendors with confidence in their product offer some form of recourse. The policy may be a full refund, a credit toward a future purchase, or a promise to re-fetch the correct document. The specifics matter less than the existence of the policy.
The catch is that many vendors define "defective" narrowly. A document that is unreadable due to poor scanning may be considered defective. A document that is the wrong case entirely may be considered a user error if the buyer clicked the wrong listing. A document that is missing pages may be considered "as archived" if the court itself was missing those pages. The buyer should read the return policy the way a lawyer reads a contract, because it is a contract.
The practical move is to email the vendor before purchasing with a specific question about the document. The response time and the quality of the answer tell a person more about the vendor than any marketing page. A vendor that responds within an hour with a clear, specific answer is a vendor that has its act together. A vendor that responds in three days with a form letter is a vendor that will be equally slow when a document arrives corrupted.
Customer support is the canary in the coal mine
The quality of a vendor's customer support is the clearest window into the quality of its product. This is true for every industry, but it is especially true for legal research, where the product is time-sensitive and errors have consequences. A vendor with a phone number that connects to a human being is a different species from a vendor that only offers a contact form that generates a ticket number.
The test is not whether the support team is friendly. The test is whether the support team can solve a problem. A person should ask a question that requires the support team to actually look at a document, such as "does page 14 contain the signature block?" If the support team can answer that question, they have access to the files and the ability to check them. If the support team responds with "all our documents are high quality," they have neither access nor ability.
The absence of a phone number is not disqualifying on its own. Many legitimate vendors operate entirely online. But the absence of any path to a human being, combined with a strict no-refund policy, is a combination that should give any reasonable buyer pause. That combination describes a business model that does not need to satisfy its customers because it does not expect them to return.
The cheapest option is rarely the least expensive one
A person who buys a defective document for five dollars has not saved five dollars. That person has spent five dollars plus the time required to discover the defect, plus the time required to find the document elsewhere, plus the risk that the deadline was missed during the detour. The true cost of a bad purchase is always higher than the sticker price, often by an order of magnitude.
The vendors that charge premium prices often justify the premium with better sourcing, better scans, and better support. This is not always true. Some premium vendors coast on their reputation while their databases rot. But the inverse is almost always true: the rock-bottom vendors are rock-bottom for a reason. A person who needs a case document for a filing tomorrow should not be shopping at the legal equivalent of a dollar store.
The final check before clicking pay is not about the document at all. It is about the buyer's own situation. Does this document need to be filed with a court, or is it for research? If it needs to be filed, the official version from the court is the only acceptable version, and no commercial vendor should be involved. If it is for research, the vendor's version is acceptable as long as it is legible and complete. Knowing which situation applies changes the entire calculus.
A lawyer who has verified the source, checked the page count, matched the caption, reviewed the docket text, confirmed the image quality, understood the delivery timeline, totaled the fees, read the return policy, and tested the customer support has done the work. That lawyer can click pay with reasonable confidence. The lawyer who skipped those steps and clicked pay anyway will eventually learn why the checklist exists. The lesson will cost more than the document.
