Guides

How to Read a SaaS Contract Before You Build on Top of a Platform


Before you write a single line of integration code, read the platform's terms of service, developer agreement, and acceptable use policy as three separate documents. Most founders read one, skim another, and ignore the third. The clause that kills your product is almost always in the one you skimmed. This guide covers the specific contractual terms that carry real risk for builders: not the boilerplate you can safely ignore, but the provisions that determine whether the product you spend six months on can actually exist.

Why This Is Different From General Legal Advice

Most legal guides for founders talk about incorporation, IP assignment, and employment contracts. Those matter. But if you are building a product whose core functionality depends on a third-party API, the most immediate legal risk you face is not in your own agreements: it is in the contract you accepted when you clicked 'I agree' on someone else's developer portal. That contract governs what you can build, how you can charge for it, who owns the data flowing through it, and under what conditions the platform can shut you down. It deserves more than a five-minute skim.

The Five Sections That Actually Matter

SaaS developer agreements tend to run long and repeat themselves. Most of the length is padding. Focus your attention on these five areas, in this order.

  1. Licence scope and restrictions: what you are actually permitted to do with API access
  2. Data rights and data portability: who owns what, and whether you can move it
  3. Termination and suspension clauses: how quickly they can cut you off, and for what
  4. Change-of-terms provisions: whether they can rewrite the rules on you mid-build
  5. Commercialisation restrictions: whether you can charge money for what you build

1. Licence Scope and Restrictions

The licence section grants you access, but the restrictions section takes most of it back. Look specifically for phrases like 'non-transferable', 'non-sublicensable', and 'personal use only'. If the licence is non-sublicensable, you cannot pass API access downstream to your own users without a separate arrangement with the platform. That matters enormously if your product is itself a layer on top of their layer. Also check whether the licence is limited to specific use cases. Some platforms grant access for 'internal business purposes' only, which technically excludes building a commercial product on top.

Red flag: any licence that grants access for 'personal' or 'internal' purposes only. If you are building a product for paying customers, you need a commercial developer licence, not a standard user agreement.

2. Data Rights and Data Portability

This is where founders get caught out badly. There are two distinct questions here and they are easy to conflate. First: who owns the data your users put into the platform via your integration? Second: can you extract that data if you need to leave? Many platforms assert a broad licence over data processed through their API, including derivative data and aggregated data. Some go further and claim ownership of any insights or outputs generated using their infrastructure. Read the data section carefully for phrases like 'royalty-free licence to use your data' or 'anonymised aggregate data'. These are legal mechanisms that let the platform monetise data your users created. If you are building for B2B customers who are sensitive about data, that clause is a commercial liability for you.

On portability: check whether the terms explicitly allow you to export data in a usable format, and whether there is a defined timeline for doing so on termination. UK builders also need to consider whether data processed through the API is subject to UK GDPR obligations. If personal data flows through a US-based platform's API and their terms make them a data processor, you need a Data Processing Agreement in place before you go live, not after.

UK GDPR requires a signed Data Processing Agreement (DPA) before any third party processes personal data on your behalf. Most large platforms have a standard DPA available on request or in their compliance portal. If you cannot locate one, treat that as a blocker.

3. Termination and Suspension Clauses

This is the clause most founders do not read until it is too late. Platforms typically reserve the right to suspend or terminate API access immediately, with no notice, for a breach of terms. The definition of 'breach' is often broad enough to include things you would not think of as violations: exceeding rate limits repeatedly, using the API in a way the platform later decides conflicts with their own roadmap, or simply operating in a market they decide to enter themselves. Look for the notice period on termination for convenience, meaning termination for no reason at all. Thirty days is common. Some agreements give as few as seven. That is not enough time to rebuild a product's core functionality.

Red flag: termination for convenience with fewer than 30 days' notice, or immediate suspension rights with no cure period. Both mean the platform can end your product before you have any realistic chance to respond.

4. Change-of-Terms Provisions

Almost every SaaS agreement reserves the right to modify terms at any time. The question is what notice they are obliged to give you before those changes take effect. Some platforms commit to 30 or 90 days' notice for material changes. Others say continued use of the API constitutes acceptance of any new terms, with as little as 14 days' notice. If pricing terms can be changed unilaterally mid-contract, your unit economics can be rewritten without your agreement. This has happened repeatedly across the developer ecosystem, particularly around API pricing. Build your financial model on the assumption that terms will change, and check what protection, if any, the agreement actually gives you.

5. Commercialisation Restrictions

Some platforms explicitly prohibit building competing products. Others prohibit reselling API access, which can affect how you package your product. A few restrict the use of platform-generated content or outputs in your own product's marketing. The most common commercialisation trap is a 'no competing products' clause buried in the acceptable use policy rather than the main agreement. If the platform's roadmap overlaps with your product category, and many do, that clause can be used to shut down your product even if you are not in direct competition today. Also check whether the agreement includes a right of first refusal or acquisition rights over products built on the platform, which is rare but not unheard of.

A Practical Reading Process

  1. Download all three documents: the main Terms of Service, the Developer Agreement or API-specific terms, and the Acceptable Use Policy. If pricing terms are separate, add those too.
  2. Search each document for these keywords: 'terminate', 'suspend', 'modify', 'change', 'compete', 'own', 'licence', 'sublicence', 'resell', 'personal data', 'aggregate'. Read every paragraph where they appear in full.
  3. Note any clause you do not fully understand and flag it specifically. Do not assume ambiguity is in your favour; in practice, it rarely is.
  4. If the product is commercially meaningful, spend one to two hours with a UK technology solicitor reviewing the flagged clauses before you commit engineering time. This is a fixed cost against an open-ended risk.
  5. Document your reading with a simple internal note: what you are permitted to do, what you are not, and what would trigger termination. Review it before every major product decision that touches the integration.

The Clauses That Are Usually Fine to Accept

Not everything in a developer agreement is a trap. Indemnification clauses that require you to defend the platform against third-party claims arising from your use are standard and reasonable. Attribution requirements, branding guidelines, and restrictions on reverse engineering are normal and unlikely to affect your product. Rate limits and throttling provisions are operational, not legal risks. The length of a terms document is not a signal of danger; it is usually just legal thoroughness. The risk lives in specific clauses, not overall length.

What to Do If the Terms Are Unacceptable

You have three realistic options if the terms carry risks you cannot accept. First, negotiate. Larger platforms increasingly offer enterprise agreements with negotiated terms for builders at meaningful scale; it is worth asking. Second, architect around the dependency. Build your product so the third-party integration is a feature, not the foundation. That means your core data model, your user accounts, and your primary value proposition should function without the API. Third, walk away before you build. The sunk cost of a week's evaluation is always cheaper than six months of engineering on a platform that then terminates your access or rewrites its pricing. The contractual risk does not go away because the product is exciting.

Do I need a solicitor to review SaaS API terms before I build?

Not for every integration, no. If the API is a minor convenience feature in your product, a careful personal read is probably sufficient. If the API is load-bearing, meaning your product cannot meaningfully function without it, then a one-off review by a UK technology solicitor is worth it. The cost is modest relative to the engineering time at risk. Look for solicitors with specific experience in technology contracts rather than general commercial practice.

What is the difference between a Terms of Service and a Developer Agreement?

The Terms of Service governs your use of the platform as a user. The Developer Agreement (sometimes called an API Agreement or Platform Policy) governs what you can build on top of it. They are separate documents and both apply to you as a builder. Many founders only read the Terms of Service and miss the restrictions that live in the Developer Agreement specifically.

Can a platform legally shut down my product with no notice?

In most cases, yes, if their terms permit it. Developer agreements are contracts you accept, and if those terms include immediate suspension rights, you have agreed to them. UK contract law does provide some protections, including implied duties of good faith in certain contexts, but these are difficult to enforce in practice against a large platform operating under US law with a US jurisdiction clause. The practical protection is reading the terms before you build, not litigation after.

What should I look for in an API agreement's data clauses as a UK-based builder?

Check three things specifically: whether the platform is acting as a data processor or a data controller for personal data flowing through the API, whether a Data Processing Agreement is available and signed, and whether the platform can use your users' data for its own purposes such as training models or building aggregate products. UK GDPR places obligations on you as the data controller, and you cannot outsource that responsibility just because the data sits on someone else's infrastructure.

Is platform dependency risk a legal issue or a product architecture issue?

Both, and they compound each other. The legal risk is what the contract allows the platform to do to you. The architecture risk is how exposed your product is if they do it. Addressing only one of them leaves you half-protected. The strongest position is a product designed so it can survive a platform change, built on a contract you have actually read and understood.