Deciding to Build Vs. Leverage

Problem

As a technical lead, I often faced the dilemma of building a solution from scratch or leveraging existing technology or framework to solve the business problem. Designing and developing platforms, services, and systems is a stimulating experience, and most engineers would love to be part of that experience. At the same time, building something new takes time, comes with various risks, and may not be the best way to solve an immediate business need. Leveraging a solution comes with the risk of potentially being suboptimal and vendor lock-ins.

In my 10+ years-long journey as a software engineer, I had to choose one over the other on multiple occasions. There is also the third choice -- build parts of the system and leverage the rest. What should be the proper structure to think and make this decision? What metric or criteria should be used to assess the feasibility of choices?

Actions taken

The first action I took was to ensure that the available solutions (from open source or vendors) met all the functional requirements of the business. The features were generally specific to the use-cases and I had to be careful about that. For instance, when we evaluated open source solutions and cloud alternatives, we compared them against the elements essential to the business.

Next, I ensured that the solution fitted well for our long-term strategy and scaled well for our business. The proposed system should work within the boundaries of our infrastructure, integrate easily with existing services, and offer the strict SLAs required by the stakeholders.

I then did a holistic analysis of the potential solutions in terms of the overall cost to the business. Operating costs are tricky, as they involve engineering, tooling, cloud, and other integrations required operations and monitoring needs. This is why I decided to connect the dots and tie them together in a comprehensive piece of analysis.

I reviewed the security, privacy, and compliance requirements of the use case to make sure we comply with the contractual and compliance requirements of our customers. GDPR compliance was a must-have feature, so I made sure the solutions are meeting that specifically.

Finally, I took additional care in understanding the support offered by vendors or open-source solutions. We created a few support tickets and evaluated the response times and the time to resolve the issue. I picked the open-source solution that had an active community and provided timely resolutions.

Lessons learned

• Building an in-house solution with your engineering teams solves your unique problem and is customized to your needs. This could be the most optimal way to solve the problem and provide a competitive advantage to the core business.
• As a technical lead, you should keep yourself up to date with the latest trends, frameworks, and technologies in your area of expertise. You should frequently reevaluate the architecture, efficiency, and cost of current solutions and evaluate potential changes or upgrades to the system.
• A deep understanding of the internal tools, platforms and infrastructure, and the partner teams supporting those is critical in recognizing what parts of the system can be leveraged.
• Prototype potential solutions against known use-cases and evaluate them on a standard set of metrics encompassing all the functional, technical, and strategic requirements.
• If you choose open-source solutions, pay close attention to the security and privacy aspects of the system and make sure the project has an active community with a quick turnaround on bug fixes and feature requests.