Knowing accessibility to code well. Don’t learn to code well to do accessibility.
Is accessibility a simple theoretical overlay to have a compliant code? Not quite. Most accessibility audits show very low scores of compliance with the RGAA (which is a not very demanding standard). These audits reveal technical problems but also design and organization problems and especially a lack of web culture.
But then, couldn’t increasing the web culture in a team also benefit performance improvement? Couldn’t correcting accessibility anomalies improve page load times?
Let’s take a step aside to see the webperf from another angle and see what accessibility could bring to it.
If you work in the world of web performance, you know that we can be confronted with complex problems. How do we speed up the experience of users with slow-loading pages, not speed up the browsing experience of those who already have the best display speed?
What if we used proven methods to solve them? As the creator of the famous “Sherlock Holmes method”, we will see how Sir Arthur Conan Doyle will help us explore a systematic diagnostic methodology applicable to web performance.
With this approach, you will be able to identify and solve technical problems with ease. Using real-world examples, you’ll learn how to collect data, analyze the results, and formulate a strategy to solve the problems.
Introduction to Sherlock’s method: observation, deduction, abduction, evaluation, verification and communication - Concrete application to web performance:
Observation: good practice of observability tools, contextualization of incidents, statistical analysis of Web Core Vitals contributors.
Deduction: good knowledge of the rendering cycle and its consequences
Presentation of several practical cases:
LCP degradation on a particular segment of the audience
Identification of user interactions
If you need to design an optimization plan, or if you monitor your site’s performance on a daily basis for the slightest regression, this presentation is for you.
As a web performance enthusiast for more than 15 years, I have acquired a solid experience as a “perf coaching” team manager at Peaksys, technical subsidiary of Cdiscount, for several years. Today, as Product Owner of an innovative web performance and security solution, my mission is to provide innovative solutions to ensure a quality user experience and enhanced security.
Outside of my job, I am also passionate about mind mapping and electronic music composition.
PO, SEO or project manager roles? Let’s put LightHouse aside and start with more precise tools to guide your developers and judge output performance. At this level, calibrating Webpagetest or Chrome DevTools and reading them carefully will allow you to have realistic project management.
Frontend developers? Let’s look at the details and manipulate these two expert tools. Bonus for heavy JS applications: tools and methods to lighten the frontend.
Lazy Hydrate, Never Hydrate, Resumable JS : frameworks’ techniques to reduce Total Blocking Time #
Total Blocking Time is the Core Web Vital metric that weighs the most in your Lighthouse score, with 30% of the final score. Whether you’re curious to understand what’s behind this metric or you’ve already busted your butt trying to optimize it, this talk is for you!
SSR rendering of front-end frameworks like React or Vue can display a pre-built HTML page to optimize the Largest Contentful Paint (LCP) and Cumulative Layout Shift (CLS). However, once this first display step is done, you still have to make the site dynamic: this is the hydration phase. This phase instantiates all the components of the page with their data and all the associated listeners.
How to measure the impact of this phase on the Total Blocking Time? Is hydration the only source of blocking time? How to concretely put into practice the recommendation to lighten the main thread? What is the answer to this heavy hydration phase proposed by the most recent front-end frameworks like Astro or Qwik?
Let’s answer these questions through concrete examples taken from 6 months of support of a high traffic e-commerce site (French top 10), with as a result a 25% improvement of their TBT and 20 points gained on their Lighthouse performance score.
Passionate about performance and scalability, Kevin has been working in web development for over 10 years. Now an architect and member of Theodo’s tactical team, he advises the group’s clients on their design or optimization needs and assists the teams in the development of public or specialized sites.
Presented in French by Martin Guillier
Martin is Engineering Manager and leader of the Performance offer at Theodo. Curious by nature, he is interested in new technologies and the improvements they bring.
You’ve read up on web performance, audited your site for ways to improve, and have a list of optimizations longer than your existing backlog. Now what? Learn how to estimate impact according to your organization’s goals.
Mel co-organizes the NY Web Performance Meetup and recently co-authored the Web Performance Chapter of Web Almanac 2022. She is currently a Senior Software Engineer II on Etsy’s Web Performance team and a Google Developer Expert in Core Web Vitals.
Noisy Real User Monitoring (RUM) data can ruin your day! We introduce a fresh concept called “Human Visible Navigations” (HVN) to tackle this risk; we focus on the experiences you actually care about when talking about the speed of our sites:
Human: We exclude noise coming from bots and synthetic measurements.
Visible: We remove any partial or fully hidden experiences. These tend to be very slow but users don’t see this slowness.
Navigations: We ignore lightning fast back-forward navigations which usually have few optimisation opportunities.
Adopting Human Visible Navigations provides you with these key benefits:
Fewer changes staying below the radar
Fewer data fluctuations
Fewer blindspots when finding bottlenecks
Better correlation with business metrics
This is supported by plenty of real world examples coming from the world’s largest scale modeling site (6M Monthly visits) in combination with aggregated data from the brand new rumarchive.com
Up to 22% of RUM data can be considered noise.
Bad data from a scraper caused a release to be wrongfully rolled back.
With HVN, Real Bounce Rate decreased by 4 percentage points.
Tim Vereecke loves speeding up websites and likes to understand the technical and business aspects of WebPerf since 15+ years. He is a Web Performance Architect at Akamai and also runs scalemates.com: the largest (and fastest!) scale modeling website on the planet.
I spent 15 years explaining to my clients why it was necessary to respect this or that good development practice. Like many, I relied on authority, citing studies conducted by this or that major company, but without necessarily having the ability to project or measure the value of the optimizations I was proposing.
I now have access to more data. And I’m eager to share with you what I’ve learned about the value of our work and why it should be crucial to embed these thoughts in all digital/marketing structures.
Boris is a Solution Expert at Contentsquare. After having spent several years helping development teams adopt Web Performance best practices, he is now in contact with Marketing and Strategy teams and shows them, with data, why Web Quality serves their objectives.
We will first discuss the impact of code download times on crawling and crawling behavior, and we will challenge some of the conventional wisdom (such as the 500ms or 200ms average time target). Then we will discuss the impact of bad rendition times in CSR on crawling and indexing, and we will explain the points to watch out for when we are at the limit (attention to the CPU load, to the resources consumed etc…). Finally we will see what is the real impact of bad Core Web Vitals on rankings (in practice: it is a nudge from Google. The “Page Experience” indicators are only used to break the tie in the rankings).
Philippe Yonnet is the founder and president of the search marketing agency Neper. He is a recognized SEO expert and an experienced speaker. He has accompanied many large accounts for twenty years in the implementation of their SEO strategy. First president of the French SEO Camp association, he is also at the origin of the Search Y international conferences. Philippe Yonnet is also a jury member of the European Search Awards and the SEO by SEO Camp Awards.
From Analytics to Advertising, Reviews to Recommendations, and more, we rely on Third-Party Tags for critical aspects of our sites. But there’s a tension between the value that third-party tags bring and the costs they impose.
Speed Matters… the longer our pages take to load the lower our visitors’ engagement is… lower page views, lower conversions, and lower revenue.
In this session Andy will share approaches and practical steps he uses to help clients reduce the impact tags have on the speed of visitors’ experience.
For good reasons the performance community tends to have a very negative view of third-party tags, but love them or hate them they’re part of our sites.
The talk aims to explore and demonstrate some the ways we can manage them and reduce their impact on site speed.
The general concepts in the talk will be useful to anyone who is involved with managing sites and third-party tags. Some of the concepts will require technical knowledge of how networks and browsers work but others will be useable by people who are adding tags using a tag managers.
The talk with explore how to choreograph tags to reduce their impact, and then cover approaches for monitoring and debugging tag performance.
Andy is a Web Performance Consultant at SpeedCurve, where he helps customers to measure and improve the speed of their sites. He stumbled into web performance in 2008 while launching an online education service and quickly ran into the challenge of delivering rich content to schools over congested networks.
Since 2012 he’s focused on web performance full time and has worked with a wide variety of organisations from retailers and publishers to financial services and FMCG brands.
The Top 10 Webperf Mistakes (do not do this at home) #
The road to hell is paved with good intentions, and so is web performance. In more than 10 years spent optimizing sites, Stéphane Rios, founder and CEO of Fasterize, has seen a lot. And as it doesn’t only happen to others, he testifies about the pitfalls in which many sites have fallen and still fall: images, lazyload, sprites, defer, preload / preconnect, inlining, etc. … everything goes! All these actions were of course motivated by the desire to improve performance, but things did not go as planned. In short, applying good practices stupidly is not enough, you need a real intelligence (not artificial) to optimize a site.
For 10 years, Stéphane RIOS was CTO of RueDuCommerce.com, leader and pioneer of French eCommerce. Obsessed with the excess byte and time wasting, he managed large technical teams while remaining a geek passionate about Lean, dev and highly available and performing web architectures.
He then created Fasterize, a startup that accelerates hundreds of ecommerce sites with its innovative cloud-based solutions. More recently, he launched his new toy, the EdgeSEO!
Webperf Expert, he has co-organized the Webperf Paris meetup and the We Love Speed event and regularly speaks on webperf and SEO topics.
The Google Core Web Vitals (CWV) team understands the amount of web performance recommendations is overwhelming and many don’t know where to start. We’ve been working on identifying the 9 key recommendations (3 per Core Web Vital), which we think will have the most impact and which we recommend sites look at first.
This talk will explain what they are, and why they are our top recommendations for 2023.
Barry Pollard is a Web Performance Developer Advocate in the Google Chrome team, working on Core Web Vitals and the Chrome User Experience Report (CrUX). He is one of the maintainers of the HTTP Archive and it’s annual Web Almanac publication. He’s also the author of HTTP/2 in Action from Manning Publications.
Understanding Cognitive Biases in Performance Measurement #
When measuring web performance, we often try to get a single number that we can trend over time. This may be the median page load time, hero image time, page speed score, or core web vitals score. But is it really that simple?
Users seldom visit just a single page on a site, so how do we account for varying performance across multiple pages? How do we tell which page’s performance impacts the overall user experience? How do various cognitive biases affect the user’s perception of our site’s performance?
As developers and data analysts, we have our own biases that affect how we look at the data and which problems we end up trying to solve. Often our measurements themselves may be affected by our confirmation bias.
This talk is targeted at individuals who want to understand the business impact of their site’s performance, and how biases in data can affect that.
In this talk, we’ll go into different biases that may affect user perception as well as our ability to measure that perception, and ways in which to identify if our data exhibits these patterns.
HTTP/3 is here, and it promises major performance benefits! Most of these we get for free, with just the flip of a switch. Some however allow or even require tweaking by (frontend) developers for optimal performance.
In this talk, we’ll look at the high-level options you as a developer have to steer HTTP/2 and 3 behaviour, and how those actually work internally. We’ll consider preloading, lazy loading and fetchpriority/priority hints. We take a critical look at keeping your core HTML/CSS smaller than 14KB in the 0-RTT era. And we’ll also discuss the new hotness of 103 Early Hints, aka ““Server Push Killer””.
Combining networking 101, protocol history, and especially practical tips, you’ll learn what’s happening under the hood and how to tune it yourself.
Robin Marx is a Web Performance Expert at Akamai Technologies. He focuses on the performance and workings of modern Web protocols like HTTP/2, HTTP/3 and QUIC and has been a contributor in the IETF QUIC working group for multiple years.
Robin often talks about web performance at international conferences, making the complex situations more insightful to the wider public. On the weekends, he likes to hit other people with longswords.
For over a decade now, the web community has been asking the question, “Are Single Page Apps faster?”
The arguments on each side are based mostly on a mixture of belief and anecdote, since we didn’t have the data to answer these questions. Until now.
In this talk, we will discuss new technologies in the web browser that allow us to measure performance metrics for Single Page Apps in the same way as for Multipage Apps. We’ll look at their functionality, how they work, and how you can use them to improve the performance of your SPA.
Yoav Weiss has been working on mobile web performance for longer than he cares to admit, on the server side as well as in browsers. He now works as part of Google’s Chrome team, helping to fix web performance once and for all.
He takes image bloat on the web as a personal insult, which is why he joined the Responsive Images Community Group and implemented the various responsive images features in Blink and WebKit. That was his gateway drug into the wonderfully complex world of browsers and standards.
When he’s not writing code or web standard proposals, he’s probably biking around the French countryside, wandering the Alps, or playing board games with his family.
[GOOGLE] Using the Speculation Rules API to improve navigations (EN) #
The Speculation Rules API is a new way for pages to improve next page navigations by telling the browser to prefetch or prerender certain URLs. This leads to faster, or even instant, page URLs. In this workshop we’ll show how to add speculation rules to a web page and show they render faster.