In 15 years of software development, I have rarely seen logging and reporting prioritized above new features. However, I have seen plenty of bad decisions and customer issues that were impacted by a lack of basic data.
Understanding how your application is performing is a universal need and getting performance data should be easy to implement.
Too often, teams try to address a lack of data by either (a) developing internal tools that are hard to use and don’t scale, (b) combining third-party tools that don’t really solve the problem, or worst of all (c) ignoring data and reporting priorities all together.
This is a solvable problem. I believe that all developers would benefit from a solution that is simple, scalable and doesn’t depend on months of effort. At my last company, we established a data practice called “Product Hacking” to understand how our platform operated without forcing the typical priority and resource trade-offs with core engineering. This practice was important to our success and agility, but we quickly learned that it required its own set of tools.
The first challenge we solved was how to quickly and easily collect data from any part of our application. We launched an internal service that allowed multiple teams to easily track anything at scale in a matter of minutes. It is clear to us that other developers and teams can benefit from a similar set of tools.
With that in mind, today we are launching the Hacker Pixel (HPX) — a simple, open source project that makes it easy for any team to measure what matters in as little as a single line of code. Track anything instantly without data engineering or prioritization discussions:
- No SDK required
- Add tracking in seconds without coordinating across multiple teams
- Works in any environment (web, mobile, native app, etc)
- Extremely flexible implementation — can be as simple as loading an image
- You own the data and the entire pipeline is under your control
- No need to build data pipelines from scratch
HPX is currently in a very early phase of development. This release configures a data pipeline in your own AWS account that captures up to 4 custom parameters and routes the data to Redshift. The tool also configures network settings so you can create simple dashboards in Amazon Quicksight or connect directly to Redshift using your favorite client.
The strength of the hacker pixel is in the simplicity and flexibility that enables powerful and wide-ranging use cases:
- Developers can instrument applications and APIs to help them measure performance and debug issues
- Product managers can track feature usage and performance without cross-team coordination or data pipeline and reporting system development
- Data scientists can track behavior within black box systems such as auction platforms or recommendation engines to optimize performance
- Data teams can manage pipeline evolution by enabling iteration and validation of requirements before development
- Business and ops teams can track performance and dig into granular data to identify trends, optimize performance or troubleshoot issues
We already put HPX to work to tackle a real world problem and added tracking to this project in about 5 minutes. The data will give us insight into how often the code used, which functions are executed, and the success or failure of the code under real-life conditions.
Based on our experience running a similar tracking tool at scale we know the potential use cases are limitless. We’re excited to put a first version out there to get your questions and feedback. Please check out the repository on GitHub and send us your feedback to firstname.lastname@example.org.
Note: We call this product a ‘Hacker Pixel’ because it was a tool we developed to quickly collect and share data, and reflects the original spirit of the term ‘Hack’. We don’t use bugs or exploits to break into computer systems and you probably shouldn’t either.