IBM uses Tugboat to facilitate cross-departmental collaboration

A migration to a new CMS platform is never easy. Throw in multiple business units, each the size of a large company, each with its own business logic and data to transfer over, and the complexity can become a tangled mess. This is precisely the challenge IBM faced. Stakeholders needed ways to test discrete chunks of code while also conducting integration testing for the entire website. 

Tugboat’s on-demand staging environments enabled teams to work together at greater speed and with greater confidence.

The challenge

Testing feature branches developed in silos

In a given moment, different teams could be deploying changes both large and small to the website’s shared GitHub repository. A quick rundown of different branches that could be open:

• Bug fixes
• Maintenance upgrades for the current platform
• Design changes
• Migration scripts
• New templates that will affect almost every page
• New fields that will affect a big chunk of editors

Developed in their own silos, each branch was difficult to test across teams. Every release required comprehensive cross-team testing during a one-week testing window. Coordination was time-consuming and fragile, performed on a single test server. Breaking changes could be introduced or a branch could be pulled before release due to inadequate testing.

The solution

On-demand environments to conduct cross-team testing

Tugboat allows teams to test code as discrete pieces of work, but also test against other features being developed by other teams. All before merging into a release branch.

Every team has unlimited environments to test their work against the last stable release, simultaneously. A team can merge multiple pull requests into a team-wide branch and test multiple changes. Finally, project managers can coordinate to pull code from other teams, allowing them to test everything across development silos. Teams can then deploy their work with confidence during a release window.

Tugboat makes communication and collaboration easy and cost-effective. Fewer surprises. Fewer release disruptions.

The result

• Bugs are identified and fixed faster — Teams correct issues incrementally and don’t have to wait for the time-crunched release window.
• ‍Greater confidence in production releases — Code is less likely to introduce regressions. QA can focus on testing against production data rather than worrying about new things breaking.