What a data licensing broker should do before earning a fee

A service description can sound complete while leaving the next move unclear. Finding buyers, helping with preparation and coordinating a deal are activities. The owner needs to know what those activities will produce: a requirement it can accept, an exclusion it can approve, or a quotation it can decide to fund. The difference becomes visible when a broad expression of interest meets a particular collection.

Consider an invented engagement. The publisher, buyer, collection, requirements, actions and decisions throughout this example are fictional; this is not a Rightmark case. The publisher's backlist contains original fault-diagnosis text alongside third-party illustrations. The buyer has asked about maintenance material for model training. The broker proposes to qualify that request and coordinate a written scope. At this point, the owner has authorized no sample, preparation budget or license.

The first useful question is whether the buyer needs the illustrations. Assume it confirms that English fault descriptions and repair instructions are enough for its proposed evaluation, provided each record includes a document identifier and section title. That answer changes the work. The parties can now consider a text subset rather than treat every item in the backlist as part of the proposed delivery.

The broker brings the publisher a scope note with three connected decisions:

The sequence matters. The buyer's answer makes a text-only proposal useful. The owner's exclusions bound the material a preparer can quote for. The quotation then makes the next expenditure reviewable. Asking a preparer to price the entire backlist before resolving those questions would be asking for a different job.

There is still substantial work with the other parties. The owner and its advisers assess authority to license the selected text. The preparer performs the approved extraction. The buyer tests whether the representative sample is useful. The broker can assemble their inputs and carry a question to an answer without becoming the expert who makes each judgment.

Nor must the answer be yes. If the buyer needs the excluded illustrations, the scope note exposes a dependency the text-only proposal cannot satisfy. The parties can investigate the relevant permissions or stop. A smaller proposed collection is useful only if it still answers the buyer's requirement; excluding difficult material does not make it a substitute for what the buyer needs.

A promise to coordinate licensing should meet this standard: the output should change what someone can decide. A qualified introduction may be valuable on its own. But when the owner has to obtain the requirements, reconcile the exclusions and commission the entire scope note, it should recognize how much of the broader service remains on its own desk.

The publisher now has a bounded candidate subset. That does not mean the broker needs possession of the archive, authority to grant rights or responsibility for extracting the files. Those are different jobs, and an engagement becomes easier to assess when the parties name them separately.

Rightmark's published buyer process describes owner-approved connections or exports, and licensing with the owner or, on code routes, a chosen licensing partner. Its unsigned Owner Engagement Agreement v4 makes another useful distinction: it separates coordination from hands-on preparation, which requires its own written scope and pricing. The draft also reserves the owner's licensing decision rather than allowing the broker to bind it. These are specific arrangements to examine, rather than a single role implied by the word broker.

Continue the fictional publisher engagement with additional assumptions. The owner approves a limited representative sample for evaluation under agreed confidentiality terms. A separate preparer will work in the owner's environment. The proposed eventual license would run directly from publisher to buyer. The broker will coordinate the handoffs, while the publisher retains the archive and authorizes disclosures.

Now give the handoff an ordinary defect. The preparer's quotation includes the text and section titles, but omits the document identifiers the buyer requested. Forwarding that quotation to the owner would pass along a document without resolving the mismatch. The broker's agreed task is to return it to the preparer, obtain a corrected deliverable description and put the correction before the owner for approval.

That small correction explains what responsibility for a handoff can mean. The preparer has a specific field to include. The owner can see whether the corrected work still fits its intended scope. The buyer can check the resulting sample against the requirement it supplied. None of that requires the broker to perform the extraction itself, and none permits it to expand the approved subset on its own.

The next handoff needs an answer too. Assume the sample is made available as authorized and the buyer reports that its evaluation requires some of the excluded illustrations after all. The broker records the changed requirement and returns it to the publisher. The owner can consider a new scope or decline. Treating the buyer's new request as permission to include the images would erase the exclusion the parties had just worked to establish.

A narrower introducing role deserves an equally clear boundary. Rightmark's Scout disclosure guidance requires Scouts to disclose that they are paid for an introduction, avoid promises of a price, buyer or approval, and keep private archive contents out of the introduction. That is a recognizable service with a limited job. It need not acquire file custody or licensing authority to be useful.

For the publisher, the distinction is practical: appoint someone to follow the missing identifier through to an approved correction, or keep that work with a named internal person. A contact who is available for questions and a coordinator responsible for returning a usable answer offer different levels of service. Neither description alone tells the owner who can sign, release files or receive the license payment.

The data licensing broker business model becomes easier to judge when the service scope sits beside the money. Who pays the commission matters, but so do the amount it is calculated on, the event that makes it payable and the work excluded from it. A fee can leave an agreed license payment intact while other expenses still reduce the owner's cash remainder.

Rightmark's published fee models distinguish standard licenses, with the buyer paying its fee on top, from premium and bespoke deals, where the owner keeps 75% to 85%, agreed in writing. The homepage also says buyers cover extraction, scrubbing and transfer. Those advertised terms should be matched to the actual proposal. The unsigned v4 engagement template accommodates an owner-paid commission and an expressly agreed buyer-funded alternative; it is not an executed universal agreement.

A Rightmark team member handling owner discussions located that choice in the particular confirmation in a November 13 proposal: “The template leaves the rate and who pays it to that confirmation”. He separately offered buyer funding for that opportunity. The practical question is therefore which arrangement the parties have accepted, rather than which payer a reader associates with brokerage in general.

To complete the publisher example, assume the parties accept the following terms after evaluation. Every amount, rate, allocation and payment outcome here is invented for the calculation, not a market price or a Rightmark quote. In particular, this example assigns preparation to the owner, separately from the public offer described above.

The buyer pays the publisher $100,000 directly for the defined license. It separately agrees to pay the broker 10% of that license price, or $10,000, when the corresponding license payment is made. The broker has no authority to receive the license money. The publisher owes no brokerage commission under these assumed terms, and buyer nonpayment does not transfer the fee to it.

The broker's scope covers the scoping, coordination and follow-through already described, including the corrected preparation brief. Carrying out that brief costs a separate $6,000, paid by the publisher to the preparer without reimbursement. Assume this is the whole preparation charge. The buyer spends another $4,000 on its own technical review and accepts a total listed outlay of $114,000.

For the arithmetic, assume both the license payment and broker fee are paid in full. Exclude taxes, currency conversion, financing, refunds, internal owner time and other unlisted costs. The preparation charge may have been incurred before the license was agreed; it is not assumed to depend on a successful transaction.

The publisher can calculate its $94,000 remainder from its own receipt and preparation charge. It does not need the buyer's internal review budget to do so. Both records are visible here because the example supplies them; a real counterparty's undisclosed costs should remain unknown. The owner's remainder is not net profit, and the broker's $10,000 is revenue before its own costs.

The distinction between coordination and execution now has a price attached. Obtaining a corrected preparation brief sits inside the broker's $10,000 service. Performing the technical work sits in the preparer's $6,000 charge. If a proposal says preparation is included, the owner needs to know which of those jobs it means. Otherwise, it cannot tell whether two quotes cover the same work.

Referral compensation can be another internal broker cost. Rightmark's Scout page says that it pays the introducing Scout from its own income. Such a payment does not automatically create an additional owner deduction. In the fictional records above, no referral amount is assumed; the allocation of the agreed license price is already specified.

There is a further distinction between the invoice payer and the effect of negotiating within a budget. Change one assumption: suppose the buyer instead sets a $100,000 ceiling for the license plus brokerage, excluding its separate $4,000 review cost. Keep the requested scope and the fee of 10% of the license price unchanged. A proposal filling that ceiling would satisfy:

License price + 10% of license price = $100,000.

That gives a license price of $90,909.09 and a fee of $9,090.91, rounded to cents. After the same $6,000 preparation charge, the publisher would retain $84,909.09 from the listed cash flows if it accepted. This is a different proposal, not a deduction from the first agreed $100,000 payment. The publisher could decline; the parties could also change the budget, fee or scope.

The calculation cannot settle whether the broker's work is worth its fee. It makes the decision possible by separating the service bought, the costs retained and the proposal actually on the table. Without those distinctions, a buyer-paid label can conceal as much as an unexplained owner percentage.

The strongest objection to buying a full-service mandate is straightforward: the parties may already have what they need. An owner with a suitable buyer relationship, an agreed scope and people available to finish the work can choose a direct license. It need not buy coordination merely because licensing involves several decisions.

Marina Micheli and her coauthors make this case in the Joint Research Centre's 2023 report, Mapping the landscape of data intermediaries. They consider an exchange that is easy and costless, for example because the parties know and trust each other and have compatible technical protocols and a suitable license. In that conditional case, “they would not need a third party for data sharing to happen”. The authors also explain how known counterparties can still face bargaining costs an intermediary helps resolve. Familiarity alone does not finish the work.

Return to the fictional publisher, but change its starting circumstances before it accepts any broker mandate. The buyer is already a repeat commercial contact. Its text requirements are confirmed, the publisher has selected the eligible subset, and the relevant rights records and suitable terms are ready for the parties' advisers. The publisher's operations lead has capacity to obtain the preparation quote and manage the sample handoff. The buyer has named the person who will evaluate it.

In that situation, the operations lead can catch the missing document identifier, obtain the corrected quote and bring it back for approval. There is no substantial unassigned coordination task for the proposed broker. Direct dealing is an ordinary, credible choice. The preparer still charges, and the operations lead's time still costs something. The case for going direct is that the parties can carry the work competently, not that doing it internally is free. Nor does avoiding a broker fee mean the buyer will add that amount to the license price.

Now change the obstacle. Suppose the publisher has the same clear scope and execution capacity but no route to a suitable buyer. A qualified introduction can be the entire contribution worth paying for. Its value need not be proportional to the minutes spent sending an email: it may create an opportunity the owner could not otherwise reach. Agreeing an introduction and compensation scope in advance can be sufficient. Requiring a useful introducer to bundle unnecessary coordination around that connection would make the service less precise.

There is also a middle case. The publisher and buyer know one another, but nobody has time to reconcile the buyer's requested fields with the preparer's deliverables. The relationship exists; the unassigned work remains. A broker offering to carry that work can be compared with assigning or hiring a coordinator. The identifier correction gives the parties a concrete task to price and evaluate instead of a general argument that intermediaries are always useful.

These are choices before accepting a mandate. They do not undo an agreed commission or justify avoiding payment for a protected introduction. The question is which contribution the owner needs to buy at the outset.

Choose the next blockage in the proposed transaction and ask the broker to describe what it will return, who can act on it and what will still need approval. One completed handoff description will reveal more about the service than a long menu of capabilities. It also gives the owner a useful comparison with its own capacity.

For the same fictional publisher, assume there is no available internal coordinator. Its review note might read as follows. This is an invented decision note, not contract wording:

Work we need covered: Reconcile the buyer's required document identifier with the preparer's deliverable description, obtain the corrected quote and return the preparation scope for our approval.

What the broker must return: The corrected description tied to the selected text and exclusions, followed by the recorded sample approval and buyer evaluation response. A request for excluded illustrations comes back to us for a new decision.

Work we retain: Supply rights evidence, approve disclosures and decide whether to license, with our advisers assessing the rights and agreement. The preparer executes the technical work; the buyer tests fit.

Costs to match: The buyer-funded broker fee covers the coordination described above. The separate preparation charge remains ours, as shown in the earlier cash records.

Decision still to make: Compare this scoped service with assigning or hiring a coordinator. The work addresses a real gap; the note does not establish that the proposed fee is good value.

The useful challenge is whether that broker task would remain if the parties already had a capable coordinator. If the answer is no, the owner has identified a direct alternative. If access is the only missing piece, it can assess an introduction service. If the handoff really is unassigned, it can negotiate a mandate that names the output and the limits of responsibility. None of those decisions requires access to confidential counterparty margins.

I would agree the contribution before treating the fee as settled. The owner should be able to point to the decision the broker will help resolve, the specialist work it must still fund and the authority it keeps. That makes the mandate a choice about useful work, rather than confidence in a broad promise to manage a deal.

Take that scope and responsibility note into Rightmark's data licensing agreement checklist to connect it with the eventual license's permissions, approvals and payment questions.