A data licence grants a use, not a dataset. The same rows can be read by your analysts, shown inside your product or handed to your customers, and each of those is a separate right with its own terms. This guide places the things teams commonly do with company data under the three uses, lists what to settle for each before you sign, and explains why we license by written agreement. Nothing in it is legal advice: use it as a list of points to raise with your own counsel.
Three uses of the same rows
What separates the three uses is where the data ends up, not how it is delivered. The table sets them side by side.
| Question | Internal use | Embedding in a product | Redistribution |
|---|---|---|---|
| Who works with the data | Your own staff, models and systems | Your product, for its users | Your customers, in their own systems |
| What leaves your organisation | Nothing from the dataset | Values and results shown in the product | The data itself, as rows, files or an API of yours |
| A typical licensee | A fund's research team, a revenue operations team | A sales, marketing or research platform | A data platform or a reseller |
| What the terms turn on | Who counts as internal | Display against export | Recipients and onward transfer |
The uses stack in practice. A platform that embeds data also studies it internally, and a redistributor usually displays it too. An agreement can grant more than one use, and it should name each.
Placing what you plan to do
The risk in a licence is the feature nobody placed: built under one use and quietly serving another. The table places common plans, with Fokals datasets as the examples. It shows where each usually falls. The agreement, not this table, decides.
| What you plan to do | Usually falls under | Why |
|---|---|---|
| Load Hiring Activity into your warehouse for analysts | Internal use | Nothing leaves your organisation |
| Train a scoring model on Company Signals and keep the scores in-house | Internal use | The output stays inside, and the model is derived data |
| Show the technologies in Technology Stack on an account page | Embedding | Users see values, in your product |
| Alert a user when a company they follow appears in Technology Changes | Embedding | A result computed from a dated record |
| Let users export enriched accounts as a file | Redistribution, unless the embedding terms allow export | Rows leave your product |
| Return the fields from an API of your own | Redistribution | A customer's system receives the data |
| Offer the datasets in your own catalogue | Redistribution | The data itself is the product |
| Print a figure from Market Series in a public report | A use to name in the agreement | Publication reaches people who are not your users |
What to settle for each use
Internal use looks simple. Its open questions are about who counts as inside and what may leave.
- Who is internal: employees only, or contractors, affiliates and other companies of your group?
- May a service provider that runs your warehouse or your models handle the data?
- May models be trained on it, and what happens to a model when the term ends?
- May a figure or a chart leave the organisation in a client note or a report?
Embedding puts data before people who are not party to the licence. The guide to enriching company records in your data product works through such a build.
- Which fields may be displayed, and to which users?
- May a user copy, export or query them in bulk? Each step moves towards redistribution.
- What must be shown beside a value: a source credit, the date it was observed, a mark on model output?
- How long may the product go on showing a value after the term ends?
Redistribution passes the data itself to your customers, under a redistribution licence. The use case on redistributing company data under licence is the working list.
- Who may receive the data, and may they pass it on?
- In what form: whole datasets, a subset, aggregates?
- What attribution travels with it?
- What may a recipient keep when its subscription, or your agreement, ends?
Terms that apply to every use
Whatever the use, five things belong in the text.
- Scope: the products and datasets, and the set of companies.
- Term and renewal, and what must be deleted at the end.
- Derived data: who may keep scores, aggregates and models built on the rows.
- Attribution owed to upstream sources.
- Versions: which label and score versions are delivered, and how much notice a change carries.
The last is already fixed in our data, which makes it easy to write down. Labels and scores are produced under named, frozen label versions, and a breaking change ships as a new version with at least 90 days' notice. Each bulk export comes as a package of files with a manifest that names its period, its label versions and its licence, so the terms of a delivery are recorded beside its files.
The record itself helps with the second and third. Every observation is dated, and daily and weekly datasets are written once, after the period closes, and never revised. A score or a model built on a given period can therefore be tied to exactly the rows that produced it, which is what a clause on derived data or on deletion needs to refer to.
Why we license by agreement
We license by written agreement, for internal use, for embedding in a product or for redistribution. Three facts explain why an agreement fits.
Delivery does not tell the uses apart. Whatever the use, the data reaches you by the same REST API and the same bulk exports, and you load it where you need it. The rows are the same rows, and they do not say whether a value may be shown to a customer. The agreement does.
Scope differs with every buyer. An agreement names the products and datasets, the set of companies, which is your own, and the use. Sample data for the companies you track is sent on request, with the data dictionary and the methodology, so the scope that goes into the agreement is one you have tested.
The terms are written for your use. Fokals data is delivered direct and licensed direct, so one agreement between you and us governs every dataset you take, and its wording can follow what you build. Data marketplace vs direct licensing compares the two routes, and the data licensing page describes embedding and redistribution for platforms.
Where a licence stops
A licence settles what you may do with the data. Whether the data serves your purpose is a matter for testing, which is what the sample is for. The rules that apply to your own use, such as those for investment research or for marketing in your market, stay yours to check. And the scope of a licence is the scope of the data: a Fokals licence covers company-level data throughout, firmographic, technographic, hiring and intent data on public and private companies.
Frequently asked questions
What is the difference between internal use and redistribution in a data licence?
Internal use keeps the data inside your organisation: your staff, models and systems work with it and nothing from the dataset is passed on. Redistribution lets you pass the data itself to other parties, usually your customers, as rows, files or an API of your own. Embedding sits between the two: values are shown inside your product, and users cannot take the data away.
Can I show licensed data to my customers inside my product?
You can show licensed data inside your product only if the licence grants embedding. A licence for internal use covers your own organisation and stops at its edge. Embedding terms usually say which fields may be displayed, to whom, with what attribution, and whether users may export them. An export of rows normally counts as redistribution, so settle the wording before the feature is built.
Who owns the scores and models built on licensed data?
The agreement decides who owns what is built on licensed data. Scores, aggregates and trained models are derived data, and licences treat them separately from the rows they were built on. Settle in writing whether you may build them, whether you may keep and use them after the term ends, and whether they may be shown or sold to others.
Why do data vendors license by written agreement?
Data vendors license by agreement when what is licensed varies from buyer to buyer: the datasets, the set of companies and, above all, the use. The same data delivered the same way can be licensed for internal use, for embedding or for redistribution, and each is a different right. Fokals licenses by written agreement for those three uses, and sends sample data for the companies you track on request before the scope is fixed.
What happens to licensed data when the agreement ends?
What happens to licensed data at the end of an agreement depends on its terms, so read them before signing. An agreement can require deletion of the data, allow it to be kept for audit, or let derived data such as scores and models survive the term. For embedded or redistributed data, it should also say what your customers may keep. Decide these points at the start, when they are easy to agree.