What is CI/CD, and does a two-person tech team need it?

CI/CD explained for small D2C teams: what continuous integration and delivery do, why even two developers benefit, and a pipeline to set up in a day.

Anadoris Digital · · 4 min read

Assembly line
Photo: Samuel Zeller / Unsplash (CC0)

In many small brands, deploying means a developer running commands from their laptop, copying files to a server, or clicking through a console. It works until that developer is on leave, uses the wrong settings, or forgets a step during a rushed fix before a sale. Then a small change becomes an outage. Manual deploys also leave no record, so when something breaks a week later, nobody can say for certain what changed or when. That uncertainty slows every investigation and makes it hard to hand work to a new developer or agency.

CI/CD replaces that with an automated, repeatable process. It sounds like something for large engineering teams, but the benefits are arguably greater for small ones, because there is less spare capacity to absorb mistakes.

What CI and CD mean

  • Continuous integration: every time code is pushed, an automated system builds it and runs checks such as tests and linting.
  • Continuous delivery: changes that pass checks are packaged and deployed to staging automatically, and to production with a click or approval.
  • Continuous deployment: a step further, where passing changes go to production without manual approval.

Most small brands do well with continuous integration plus continuous delivery: automatic checks and staging deploys, with a person approving production releases. That approval step is a good moment to confirm the change is not landing in the middle of a campaign, and that someone is around to watch it go live.

Why it matters for a small team

  • Deploys work the same way whoever runs them, so knowledge is not stuck in one head.
  • Broken code is caught before it reaches customers.
  • Every release is recorded, so you can see what changed when something goes wrong.
  • Rolling back is a known, quick step rather than a scramble.
  • Secrets live in the pipeline's secure settings, not on laptops.

A simple pipeline you can set up in a day

Popular options include GitHub Actions, GitLab CI/CD, Bitbucket Pipelines and AWS CodePipeline. For a small team already on GitHub, Actions is usually the easiest. A first pipeline need not be complex.

  • On every push: install dependencies, run tests and type checks, build the app.
  • On merge to the main branch: deploy automatically to staging.
  • On approval: deploy the same build to production.
  • After deploy: run a quick smoke test against key pages or endpoints.
  • On failure: notify the team chat and stop.

Security for your pipeline

A pipeline that can deploy to production is a powerful thing, so protect it. Use short-lived credentials, such as GitHub's OpenID Connect integration with AWS, rather than long-lived access keys stored as secrets. Give the deploy role only the permissions it needs. Require reviews before merging to the main branch, and limit who can approve production deploys. Review third-party actions or plugins before adding them, as they run with your pipeline's access. Pin them to specific versions so an update you did not review cannot change your build. Keep the pipeline configuration in the same repository as the code, so changes to how you deploy are reviewed just like changes to the app itself, and anyone can see exactly how a release reaches customers.

What about Shopify themes?

Shopify themes can also be managed in Git with Shopify's GitHub integration or its command-line tool. Theme changes then go through review and can be rolled back. It is worth doing if more than one person edits the theme or if an agency makes changes alongside your team.

Pipeline services charge for build minutes beyond any free allowance. Check current pricing for the service you choose, and keep builds efficient by caching dependencies.

Database changes in the pipeline

Code is easy to roll back; database changes are not. If a release adds a column, renames a field or changes how orders are stored, run those changes through the pipeline as migration scripts, apply them to staging first, and take a backup or confirm point-in-time recovery before applying them to production. Prefer changes that work with both the old and new versions of the app, so a rollback does not break.

For busy databases, avoid running long migrations during sale hours. Changes that lock large tables can block order processing until they finish. Schedule them for quiet periods, set sensible timeouts, and have a plan to stop them if they run longer than expected. This is one of the places where a pipeline's discipline saves the most pain.

Where to start

Pick your most frequently changed app. Write down the exact steps used to deploy it today, then turn those steps into a pipeline that runs tests on every push and deploys to staging on merge. Add a manual approval step for production. Once it has run a few releases without trouble, apply the same pattern to your other apps and retire laptop deploys.

Related service: AWS & Cloud Management · More on this topic: AWS & Cloud Management articles

Common questions

Is CI/CD worth it for one developer?

Usually yes. Even a single developer benefits from automatic tests, consistent deploys and a release history. It also means someone else can deploy or roll back if the developer is unavailable, which reduces the business's dependence on one person.

Which CI/CD tool should a small brand use?

Use the one built into where your code already lives: GitHub Actions for GitHub, GitLab CI/CD for GitLab, Bitbucket Pipelines for Bitbucket. AWS CodePipeline suits teams who want everything inside AWS. All are capable for small teams.

Does CI/CD need automated tests?

It works better with them, but you can start without many. Begin with a build step and basic checks such as type checking, then add tests for the most important flows, like order creation and webhook processing, over time.

How do I roll back a bad deploy?

With CI/CD, rolling back usually means redeploying the previous successful build or reverting the change and letting the pipeline run. Practise it once on staging so the steps are familiar before you need them in production.

Your next chapter

Let’s build what comes next.

A new website. A better system. A simpler way to work. Tell us where you want to begin.

Discuss your project