Digital income
Data and API Licensing
Selling access to a dataset or an interface, priced by seat, by call, or by contract.
How the money actually reaches you
A customer licenses access rather than buying a copy. Self-serve customers pay by card on a plan with a call quota or usage meter, through a processor or an API marketplace that takes a share; larger customers sign an annual contract with a negotiated licence and pay on invoice terms, which means the cash arrives well after the service is delivered and the contract governs everything. What you are actually selling is the right to use the data in defined ways — internally, in a product, redistributed, or to train a model — and each of those is a different price and a different clause. The data itself must be maintained continuously or the licence stops being worth anything.
The customer, directly. A payment processor or a marketplace stands between you and the money and takes a fee for it; an app store or platform takes a commission set in its developer terms. What you actually sell may be a file, a licence, access, a physical parcel someone else ships, or the right to call an interface.
Fastest of the three. Card money typically settles to the processor within days and is paid out on a rolling schedule, minus a reserve on new or high-risk accounts. Marketplaces and app stores pay monthly, in arrears, after their own hold period.
Read the payer, the intermediary and the schedule before anything else. Who pays, who sits in the middle taking a share, and how long after the work the money arrives explain most of the difference between these fourteen businesses.
The structural facts
Six lines, in two halves. The first three are what the model takes — money up front, waiting time, and the rhythm of work it never stops needing. The last three are what happens once it earns: how the income fades if nothing new is added, what stops it growing past a point, and what, if anything, stops the next person copying it. A model with no answer to the last two can still pay; it just does not compound.
- Capital to start
- Moderate to high: acquiring or generating the data, building and running the interface, and the legal work to license it properly.
- Time to first dollar
- Months. Self-serve customers can sign up quickly once it exists; contract customers run procurement and legal review first.
- Ongoing effort
- Continuous. Stale data is worthless data, and customers integrating an API expect stability, versioning and uptime.
- How it decays
- Immediate. A dataset that is not refreshed is a liability rather than an asset, because customers have built systems that depend on it being current.
- What caps it
- How many organisations need this specific data and are permitted to buy it. Often a small number of large customers, which is concentration risk by construction.
- What stops a copycat
- Provenance and exclusivity: data you generate, collect under agreements others cannot replicate, or clean far better than anyone else. Data scraped from a source anyone can reach has no moat and a legal problem attached.
What it costs to run
- Data acquisition, collection or licensing in
- Infrastructure to store, process and serve it
- Engineering for the API, documentation and versioning
- Legal drafting of licence terms and data-protection review
- Support for integrating customers
How it typically fails
- The upstream source cuts off access and the product has no inputs.
- The data was collected in a way the source's terms prohibit, and the letter arrives.
- The dataset covers personal data and the privacy obligations were never addressed.
- One or two customers are the whole business, and one of them builds it in-house.
- Maintenance lapses, the data goes stale, and integrated customers leave all at once.
Does it sell?
Whether a business can be sold is the plainest test of whether it is an asset or a job. A buyer pays for income that continues without the person who built it, which means verifiable revenue, a customer relationship that transfers, and a distribution channel that does not walk out of the door with the founder. This section describes whether that market exists for this model — not what anything is worth, which no honest public source can give.
Data businesses are bought, usually by strategic acquirers who want the dataset inside their own product rather than by the individual buyers who purchase content sites. Diligence concentrates on rights: whether you actually have the right to license the data, whether the contracts survive a change of control, and whether the collection method is defensible. Unclear provenance is the thing that kills these deals.
Where people look
A storefront for selling digital files, courses and memberships, handling checkout, file delivery and sales tax on the seller's behalf.
Fee is taken per sale rather than as a monthly subscription
Visit Gumroad ↗A publishing platform for email newsletters with paid subscriptions, payment processing and a subscriber list built in.
Takes a percentage of paid subscription revenue plus card processing
Visit Substack ↗Amazon's self-publishing service for ebooks and print-on-demand paperbacks, which pays the author a royalty on each sale.
Royalty rate depends on list price band and file delivery size
Visit Amazon Kindle Direct Publishing ↗Hosted software for building and selling online courses, including payments, student accounts and drip release of lessons.
Visit Teachable ↗Sponsored links are labelled. Listing a service is not an endorsement of it, and nothing on this page is advice.
Frequently asked
Can I license data I collected from public websites?
Why do data licences distinguish between uses?
Is this passive?
Others running on the same engine
Subscription Websites and Membership Sites
Gated content or a community that members pay for monthly or annually.
Paid Newsletters
Writing delivered to an inbox, paid for by subscribers, sponsors, or both.
Mobile Apps
Software distributed through app stores, paid for by users, subscriptions or advertising.
SaaS Businesses
Software run as a service, billed monthly or annually to the people using it.
Online Courses
Recorded teaching sold once and delivered many times, direct or through a marketplace.
Digital Templates and Downloads
Files sold repeatedly — templates, presets, fonts, printables, plugins, spreadsheets.
Stock Photography Portfolios
Images, video clips and audio licensed repeatedly through agencies, paid per download.
E-Commerce With Outsourced Fulfilment
Selling physical goods someone else stores and ships — third-party logistics, print-on-demand, or dropshipping.
Domain-Name Portfolios
Holding registered domain names for lease, parking revenue or eventual resale.
Research only. This page describes how a business model works, what it costs and how it fails. It does not recommend starting one, does not estimate what anyone earns, and is not investment, tax or legal advice.