Skift IQ is my first commercial product (most of its content is currently under a trial pay wall), and so building it required me to prioritize issues of software engineering and product development that I hadn’t previously focused on before. Because news organizations traditionally operate under a “publish and move on” rhythm, there isn’t much incentive or time to implement test-driven development or automated systems deployment, nor have I had in news projects any reason to build a custom payments and accounts system.

/images/posts/skift-iq-delta-airlines.small.png

In addition to these requirements, I applied and expanded on my past development experience. We had a design firm create the look of the site, but I implemented most of the front-end code, including the JavaScript data visualizations. And I also wrote all the code needed in between to connect the visuals to the underlying data, such as efficient scrapers and queries to collect and analyze the data without crashing our MySQL server.

I’ve found that no matter the business goals, the conceptual design and programming remain nearly the same. Attractive graphs still have to show interesting data stories. Writing maintainable code requires the same abstraction of real-world concepts and processes no matter what the actual topic . And massive project goals, whether it’s for business or for the public interest, needs the best tools and creative problem solving skills.