Royalty & intellectual-property income
Content Licensing
You own articles, video, archives or data and license the right to republish or use them, charging per use, per period or as a share of the licensee's revenue.
Content licensing is income from granting others the right to republish, redistribute or otherwise use written work, video, archives or structured data you own. Deals range from syndication and white-label feeds to corporate republication licences, API and data licences, and agreements permitting use in machine-learning training. Payment can be a flat annual fee, a per-use charge, a revenue share or a minimum guarantee against usage, and the licensor can only grant what it actually owns, which makes upstream rights clearance the decisive detail.
Royalties and licensing Semi-passive
The labels above place this income type before you read a word. The first is the mechanism — how the money actually reaches you, whether by lending it out, owning a slice of something, renting an asset, licensing a right, selling an option or owning a business somebody else runs. The second says whether the income keeps arriving on its own once it is running, or whether it needs work from you to keep coming. Both are descriptions of how the thing is built, not verdicts on it.
How it works
A licence is a contract that defines exactly what may be used, in what medium, in what territory and language, for how long, and whether the grant is exclusive to that one licensee or open to others simultaneously. Syndication places existing articles or video with other outlets under this kind of grant, priced as a flat fee per outlet, a per-piece rate, or a share of the advertising revenue the piece generates once republished. White-label and feed deals go further, delivering content programmatically into a partner's own branded product, typically priced as a subscription with tiers based on volume.
Archive licensing turns a back catalogue that costs nothing further to produce into a saleable asset, and for an established publisher it is often the highest-margin line in the business. Data and API licensing works on the same logic but sells structured information rather than prose, with usage limits — call volume, seats, redistribution rights — written into the contract rather than enforced by the technology itself.
Corporate republication at scale is usually handled through collective licensing organisations, which grant blanket rights to companies wanting to circulate news or research internally and distribute the resulting fees back to registered rights holders. Licensing for machine-learning training is a newer, distinct deal type: a fixed fee for a defined corpus and term, with contractual representations about the licensor's right to grant that use.
None of this works if the licensor does not actually hold the rights being sold. Freelance contributor agreements, embedded images, music beds and interview releases all have to permit the onward grant, or those elements must be stripped before a deal closes. Search and attribution terms — canonical tags, no-index requirements, link-back obligations — determine whether syndication cannibalises the licensor's own audience, and machine-readable access controls on a website are usually the first practical step in getting unlicensed use to the negotiating table at all.
What it pays
Payment is quoted as a flat annual or monthly licence fee, a per-article or per-use rate, a revenue share on advertising the licensed content generates, or a minimum guarantee against usage with overage charges once that floor is exceeded. What any of those numbers actually come to is driven by the exclusivity granted, the size and reach of the licensee, the volume and freshness of the content on offer, the depth of the archive behind it, and how easily a buyer could get equivalent material somewhere else.
Data licences typically price on usage metrics — API calls, records returned, seats provisioned — so the income scales with the customer's own product growth rather than with anything the licensor does after signing. Non-exclusive licensing lets the same corpus be sold to multiple buyers at once, which is the point at which content licensing genuinely behaves like a royalty stream rather than a one-time sale. Exclusive deals pay more per contract but foreclose every other buyer for the term of the agreement.
An archive is a stock, not a flow: it can be relicensed repeatedly at no further production cost, but its value does not grow unless new material is added to it.
Costs and taxes
The main hidden cost is rights clearance — auditing contributor agreements, replacing images or music that were never properly licensed, and documenting a clean chain of title before any deal can close. Technical delivery adds ongoing expense too: feeds, APIs, metadata standards and usage reporting all require engineering work that scales with the number of active licensees, and legal negotiation, indemnity exposure and monitoring for out-of-scope use are recurring line items rather than one-off setup costs.
In the US, licence fees are ordinary income — reported as business income by an operating publisher or on Schedule E by a passive rights holder. Cross-border deals raise the same royalty-versus-sale characterisation question that comes up in software licensing, and that characterisation determines whether foreign withholding applies and at what treaty rate.
Payments routed through a collective licensing organisation arrive net of that organisation's administrative deduction, so the headline usage figures are not what lands in the account. State sales tax may also apply to certain digital products and information services, depending on the state and how the deliverable is legally characterised.
Liquidity and time commitment
Archives and content libraries change hands privately, priced off demonstrable licensing revenue and the breadth of rights that have actually been cleared for resale. There is no public market quoting a price per article or per dataset. Individual licence agreements are typically annual or multi-year with renewal terms, so income is contracted in advance rather than earned spot-by-spot.
Ongoing effort is real even after a deal is signed: delivery, customer support, usage reporting, renewal negotiation and enforcement against use that has drifted outside the licensed scope all fall to the licensor. Sourcing new deals requires active business development in a market that is relationship-driven rather than listed anywhere. The archive itself is not a static asset either — it needs storage, format migration and metadata upkeep to remain licensable at all.
How it goes wrong
The most consequential failure is granting rights the licensor never actually held — a freelancer's contract never assigned them, or an embedded image was used without clearance — at which point the deal's indemnity clause turns a revenue line into a liability. A related failure is structural rather than legal: syndicated copies can outrank the original in search results, and the licensor ends up losing the direct traffic it was trying to monetise in the first place.
Exclusivity carries its own risk — locking a corpus to one buyer that later loses interest still means waiting out the full term before anything can be sold elsewhere. Pricing too wide and too cheap has a similar effect over time, devaluing the content and undercutting what any premium buyer is willing to pay on the next deal. A licensee that exceeds the agreed scope — more seats, more territories, onward redistribution to third parties — is often hard to catch, since auditing actual usage on a data feed requires rights that are rarely built into the contract or exercised in practice.
Content also gets used with no licence at all, and enforcement frequently costs more than the original deal was worth, making it uneconomic to pursue. A fixed-fee training-corpus licence leaves the licensor with no recurring participation in whatever value the buyer creates downstream. And ownership itself can be contested years later, when a contributor whose original agreement was silent about future media formats claims rights the licensor believed it already held.
What to remember
- Income comes from licence fees, per-use rates, revenue shares, or minimum guarantees, and the rate is set by exclusivity, reach, volume and how replaceable the content is.
- The licensor can only sell rights it actually holds, so contributor agreements, embedded media and interview releases have to permit onward licensing or be stripped out first.
- Non-exclusive deals let the same content be sold repeatedly, which is where this most resembles a royalty; exclusive deals pay more but foreclose every other buyer for the term.
- Syndication can cannibalise the licensor's own traffic if syndicated copies outrank the original, making attribution and search terms a commercial issue, not just a legal one.
- Deals are private, relationship-driven and multi-year, with ongoing work in delivery, reporting, renewal and enforcement rather than passive collection of a check.
- Unlicensed use and scope creep by licensees are common failure modes, and enforcement is often uneconomic relative to what was actually taken.
This page explains how the income type works, which does not change from week to week, so it deliberately carries no rate and no price. The links below go to the pages that hold the current figures for it, each one stamped with the date the data was pulled. Read the mechanism here first: the numbers there are far easier to judge once you know what they are measuring.
See the live numbers: Royalties, Digital Income.
Frequently asked
What is the difference between syndication and a content licence?
Why does rights clearance matter so much?
Does syndicating content hurt the original publisher's search traffic?
How does licensing content for AI training differ from ordinary licensing?
How is content licensing income taxed in the US?
Written for information only. Nothing here is investment, tax or legal advice, and no page on this site recommends buying or selling anything. Rules and tax treatment change; verify anything that matters with a professional who knows your situation. This explainer was drafted by a language model (claude-sonnet-5) from an editor-approved outline and fact sheet, under the rules set out in our editorial policy, and carries no market figures. Last updated Jul 29, 2026.