A data sale takes as long as its slowest stage, and that is usually the rights review or the seller's own legal review. Duration depends on how many contracts must be read, how organized the records are and how much free text needs cleaning. Outrich Data does not promise a number, because the drivers are unknown before review.
Why there is no single answer
Sellers reasonably want to know how long a sale will take. Any honest answer has to start with the fact that most of the time is spent on things that vary from one company to the next: how many customer contracts exist, how they are worded, where the data lives and how clean it is.
A deal has a critical path. Some stages can run side by side, and some cannot begin until another ends. The total is set by the slowest chain, not by the sum of everything.
This guide describes each stage and what makes it shorter or longer. It gives no durations of its own. Where other buyers publish timing claims, those are quoted as theirs.
It helps to separate elapsed time from effort. A stage can involve a few hours of work and still take a long time on the calendar, because it waits on a person, a document or a meeting. Most delay in a data sale is waiting of this kind, and most of it is on the seller's side, where the documents and the decisions are.
The stages and their drivers
A sale to a direct buyer runs through the stages in the table. Qualification and scoping are quick when the seller knows its systems. The middle stages carry the weight.
Some of these overlap. A seller can begin exporting data while contracts are being read, and its counsel can review a draft license before the scope is final. Others are strictly ordered. The offer depends on the rights review, and delivery depends on escrow being funded.
| Stage | What happens | What drives the time |
|---|---|---|
| Qualification | Basic fit is checked: size, history, data types, obvious blockers | How readily the seller can describe its systems |
| Rights review | Contracts, notices and prior licenses are read against the proposed dataset | Number of contract versions; how findable they are; how many clients negotiated custom terms |
| Scope and offer | A dataset, license scope and price are set out in writing | How much the rights review excluded; whether exclusivity is in question |
| Seller's legal review | The seller's counsel reviews the license and the rights findings | Counsel's availability and familiarity; board or lender approvals needed |
| Export and de-identification | Data is extracted, cleaned and packaged for approval | Number of systems; export limits; share of free text; rounds of sample review |
| Approval and escrow | Seller approves the package; funds are placed in escrow | Internal sign-off steps; completion of escrow instructions |
| Delivery and acceptance | Files are delivered and checked against the agreed criteria | How testable the criteria are; whether any check fails and needs a cure |
Rights review
The rights review is often the longest stage and the least predictable. It depends on reading documents that the seller has to find first.
A company with one standard customer agreement, used unchanged for years, can be reviewed quickly. A company with several generations of terms, a long tail of negotiated enterprise contracts and customers inherited through acquisitions takes much longer. So does one whose signed contracts are scattered across email and personal drives.
Privacy notices add a second strand. Each historical version has to be matched to the period of data it governed. Employee agreements and contractor assignments add a third. Gaps in any of these do not always stop a deal, but each one has to be resolved or the affected data excluded.
The review can also turn up something that needs fixing before a deal can proceed, such as a missing contractor assignment or a client whose consent is worth seeking. Whether to fix it or exclude the affected data is the seller's choice. Fixing takes longer and keeps more of the dataset.
The seller's legal review
Once an offer and a draft license exist, the seller's own counsel reviews them. This stage belongs to the seller, and the buyer cannot shorten it much.
The main drivers are who the reviewer is and what approvals the company needs. A general counsel who has handled licensing before will move faster than outside counsel seeing a data license for the first time. If the decision needs board approval, the board's calendar sets the pace. If a lender holds a lien over the company's intellectual property, its consent is another step.
Negotiation of the license itself varies. Warranties, liability caps and exclusivity draw the most discussion. A buyer that states its position on these early removes some of the back and forth.
Export and de-identification
Getting data out of systems is slower than people expect. Platforms limit export speed, some require a support request, and older systems may have no clean export at all. Each additional system adds its own extraction task.
De-identification time depends mostly on the kind of data. Structured tables with known columns are processed quickly. Free text, such as messages and ticket bodies, needs automated scrubbing followed by manual sample review, and the cycle repeats until the misses are acceptably rare. Audio is slower again.
The seller then reviews and approves the package. That step should not be rushed, since it is the seller's last look before anything is delivered.
Variety matters more than volume. A very large table from one system is a single job. A modest amount of data spread across a ticketing system, a chat platform, a document store and a code host is four jobs, each with its own format and its own cleaning rules.
Acceptance and payment
Delivery itself is fast. What follows depends on the acceptance criteria. Criteria written as testable checks, such as record counts, schema, date range and a de-identification sample, can be run promptly. Vague criteria invite argument.
If a check fails, the agreement's cure period applies and the seller corrects and redelivers. Payment timing then depends on the structure. Where funds were placed in escrow before delivery, release follows acceptance under the escrow instructions. Where a buyer pays on invoice terms, the seller waits for the payment period to run.
Recurring deliveries follow a shorter path. Once the rights review is done and the export and cleaning steps are set up, a refresh under the same license repeats only the later stages.
What other buyers publish
Some buyers of company data state timing on their own websites. These are their claims, read on their pages, and they measure different things. micro1 states that it pays in full within 30 days of acceptance. Polyshares states net 30 to 60 days. Mercor states that a wire typically follows 2 to 4 weeks after signing. DataPaid states typically 20 to 60 days from discovery to payment.
Scale AI, Mode and Troveo do not state a payment timeline on the pages read. Note the starting points. Days from acceptance, weeks from signing and days from discovery are three different clocks. Only one of the four claims covers the whole process. None says how often the stated timing is met.
A seller comparing buyers on speed should ask each one the same thing: from which event the clock starts, what has to be true for it to start, and what happens if the stated period is missed. A short payment period that begins only at acceptance says nothing about how long acceptance takes.
Why no number is promised here
Outrich Data does not publish a timeline. The largest drivers are the seller's contracts and the state of its data, and neither is known until the review has started. A number given before that would be a guess, and a guess presented as a commitment is the wrong way to begin. What is committed in writing is the order of events: rights are reviewed first, the offer has a stated validity period, the seller approves the package, and payment is funded into escrow before delivery.
A seller can shorten its own side of the process with the following preparation.
- Gather every version of the standard customer agreement and the list of clients on negotiated terms.
- Collect historical privacy notices with the dates each was in force.
- List the systems that hold the data, their administrators and their export options.
- Name one senior person who owns the decision and can approve the package.
- Brief counsel early, before an offer arrives.
- Find out whether board or lender consent will be needed.
- Agree acceptance criteria that can be tested, to avoid a dispute at the end.