How to Conduct an Effective Technical Spike as a Software Engineer to Get the Answers You Need

Technical spikes are used in Software Engineering for tickets and bugs, where the solution is not yet clear.

But do you know how to best conduct a spike?

It’s easy to assume a spike is just research, but a poorly executed spike can make the implementation of a feature challenging.

Here’s how to conduct a spike to get the best results.

1. Create a scope for the spike

A spike involves research, but you need to define what that research will involve, otherwise you will end up in an out-of-focus rabbit hole.

Before you go researching a solution, you need to:

• Set a timebox — Define the amount of time you will spend on the spike.
• Create a list of questions you need to answer — You are researching because you do not have all the answers. Write down your unanswered questions.
• Form a list of requirements — Normally generated by your product manager. Requirements explain how the feature should work.

Defining the scope with the information above helps you stay hyper-focused on the task at hand.

2. Research the problem

Now that the spike has begun, you want to gather as much information as possible.

Information can be found in many places. Here are a few suggestions as to where to look.

Research online for similar problems and identify solutions

GitHub, StackOverflow, and Medium are all great places to find this.

Make sure to stay focused on the technology you are implementing, and ensure the information you are reading is up-to-date. Implementation strategies get deprecated over time, so do some due diligence.

Cross-check approaches to ensure that suggested solutions are okay to use from a security and scalability approach.

Read documentation

The best solutions are not always found in forums.