What is an API? A plain-English guide for managers
"We need to connect to their API." "The API is down." "We hit the API limit." APIs come up in almost every software project, and they often explain why a simple-sounding feature takes longer than expected.
The simple definition
An API is a set of rules that lets two pieces of software talk to each other. API stands for Application Programming Interface.
A helpful comparison is a waiter in a restaurant. You do not walk into the kitchen. You give your order to the waiter, who takes it to the kitchen and brings back your food. An API does the same job between systems: one system sends a request, the API passes it on and returns the answer.
Everyday examples
- Payments. Your checkout sends card details to a payment provider such as Stripe or PayPal through their API, and gets back "approved" or "declined".
- Maps. A store finder shows a map by asking a mapping service's API for it.
- Sign in with Google. Your app asks Google's API to confirm who the user is.
- Your own product. Your app's screens often get their data from your own backend through an internal API.
Why APIs matter for your plans
Connecting to another company's API means relying on something your team does not control. That brings a few common issues:
- Documentation. Some APIs are well explained. Others are not, and developers have to experiment to learn how they behave.
- Access. Many APIs need an account, an approval process or an API key, a secret code that identifies your app. Getting access can take days.
- Limits. Most APIs cap how many requests you can make, known as a rate limit. Exceed it and requests fail or slow down.
- Cost. Some APIs charge per request. Usage that grows with your users can become a real expense.
- Changes. The provider can update or retire their API, and your team then has to update the connection.
When an estimate for an integration looks large, these are often the reasons.
The phrases, translated
| They say | It means |
|---|---|
| "Their API is down" | The other company's system is not responding, so our feature that depends on it fails |
| "We hit the rate limit" | We made too many requests and the provider is refusing more for now |
| "We need an API key" | We need credentials from the provider before we can connect |
| "The endpoint changed" | The provider moved or changed part of their API, so our connection needs updating |
| "We will build a public API" | Other companies or customers will be able to connect their software to ours |
Questions worth asking
- "Do we already have access, or do we need to apply?"
- "Are there limits or costs we should plan for as usage grows?"
- "What happens to our product if their API goes down?"
- "Have we used this API before, or is this new to the team?"
Bring it into the conversation
Now when your team mentions an API, you know it means a dependency on another system, with access, limits and costs to plan for. For any other technical term, check our tech glossary or paste the message into Plainlee.