Stefan Judis is the creator of Performance Tooling Today, a website that is dedicated to the collection of performance-related resources. Stefan is also involved in other performance and front-end related activities as he is the organizer of the Web Performance Group Berlin, is an occasional teacher and organizer of CSSclasses Berlin, and is a Developer Evangelist at Contentful.
Today, we interview Stefan and ask him 8 questions all revolving around the topic of either web performance or web development. Check out his answers below to get more insight into what his thoughts are on the subject.
Question 1: Tell us a little bit about your background and how you’re involved in the web performance industry.
Six years ago I changed my career from being a sound engineer to become a web developer. I discovered that I like web standards. The fact that there are always things to improve when building websites got me really excited. These days my favorite topics are web performance and accessibility.
Due to the fact that there are so many resources out there, I started collecting performance-related resources at perf-tooling.today . The idea is to share all these resources and make them discoverable for everyone. Perf-tooling quickly became my go-to resource when I was looking for material. It grew to a project including hundreds of tools, talks, and articles. I add new resources whenever I come across something valuable, and I’m happy that the community values this and adds materials, too.
Perf-tooling is rather Frontend-focused right now. This is where my personal expertise is, but I’d love to see more contributions by people dealing with the server-side.
Question 2: Although it may vary by niche/industry, what do you believe is a good benchmark that webmasters should aim to achieve when it comes to their website’s load time? Additionally, can you provide a couple of “pain points” that users should prioritize when wanting to improve their website’s performance?
Load time is not a good metric anymore because it doesn’t say anything about user experience. The websites we build these days are often closer to full-fledged applications rather than content providing documents. In my opinion, user-centric metrics are the way to go. WebPagetest provides such a metric – the Speed Index. The Speed Index gives us information about the moment when most of the web page is rendered. This is already way better.
Additional to the Speed Index custom metrics can be valuable. If a given feature is the core of your business, it may make sense to track when this feature becomes usable.
- Bringing down server response time
- Minify text based files (.html, .svg, .css)
- Optimize images
- Enable compression (e.g. gzip)
- Set proper caching headers
- Reduce the number of render-blocking requests
When these things look good, it’s time to get into more advanced topics. ;)
Question 3: What are a few reasons you believe web performance is important?
For me, web performance these days is just a tiny piece of the overall picture. The goal should be to make the internet a welcoming and accessible place. Shipping websites with more than ten megabytes page weight or showing a meaningful paint after tens of seconds is not the thing I want to be responsible for, as a professional web developer.
We never know what the circumstances of our users are. It could be that they’re paying for every megabyte, that they’re using a five-year-old device or that they’re on a shaky mobile connection. These are things we just don’t know, and that’s why we should give our best to make the web as accessible as possible because the internet should be for everyone.
Question 4: What are some use-case examples of where it might be beneficial to sacrifice a bit of speed in return for features?
Well… that’s a tricky one. I guess this pretty much depends on the product you build. Core features are surely crucial for the success of it. The advice I can give here is to bring in performance monitoring with tools like Calibre or SpeedCurve as early as possible. When you have access to the right metrics, it’s easier to understand what an improvement in performance means for your users and your business.
As much I’d like to say, that we should always go “performance-first”, in the end it’s a business…
Question 5: Please provide a list of 3-5 of your favorite tools to use when working on improving a website’s speed (this can be performance test tools, caching tools, minification tools, etc.).
My favorite tools are testing tools because without them it would be hard to improve web performance:
Question 6: What are some essential technical skills someone should know if they want to run a successful website?
I feel like a lot of people lack a solid understanding of Frontend technology these days. In 2012 Steve Souders said that 80-90% of the end-user response time is spent in the Frontend, and this is still true!
Knowing how a browser works can play a significant role to run a successful website. Because without it optimizing can be hard. You want to know when and what is slowing down the user experience.
Additionally keeping track of latest technologies is also important. HTTP/2 can be a low hanging fruit and offers a lot of improvements which shouldn’t be missed.
Question 7: Keeping up to date with the latest web development and performance standards is important for people in the web/tech industry. What are a few of your favorite performance and web development blogs or resources?
I usually try to keep track with everything web development related.
To do so, I follow people like Jake Archibald, Paul Lewis and Addy Osmani on Twitter. These folks do an excellent job educating the world about Frontend performance.
Apart from the Twittersphere, I’m subscribed to the weekly newsletters by Cooper Press. These are an excellent resource to stay up to date.
Question 8: Do you see any technical trends in website development and/or performance? Additionally, do you predict any changes to the web landscape in the next 5 years?
Right now I see two trends happening.
This first trend goes towards these closed, controlled environments. Facebook’s Instant Articles, Google’s AMP and Apple News are examples of the big companies. These “platforms” try to guarantee a great user experience including an excellent performance by limiting the developers to a subset of functionality and controlling the environment.
The reasoning behind this is that’s it simply too easy to worsen performance and degrade user experience. Unfortunately, the overall state of the web proves them right.
Personally, I’m a bit worried about these movements because controlled environments go against the ideas of an open web, but let’s see where the journey goes.
The second trend for me is progressive web apps (PWA). Offline functionality, controllable caching and the possibility to build app-like experiences is exciting. A lot of people think that progressive web apps are only about making websites offline capable, but PWAs aim for an excellent user experience and performance plays a significant role in that. This might change how we build things for the web.