Web Performance Calendar

Working From Home

Working from home has been an adventure for most folk. What’s been interesting to watch is the increase in web activity globally. I have been fortunate enough to interact with people in different countries, and while discussing with them, I noticed the frustration they face when attempting to access content on the web.

Web Frustration

One of the countries on my radar is Nigeria. With approximately 200 million people and over 170 million mobile devices, this was an excellent study to gauge user frustration on the web.

Gathering information about the top 1000 websites visited in Nigeria (CRuX), I noticed that 33% of the websites had an LCP score of over 5 seconds for their origins.

I can imagine the regular impatient 21st-century internet user leaving the page for something else.

Top 1000 websites	Top 100000 websites

Connection Types gathered from the 1000 websites and 100000 websites (CRuX), November 2021

The charts above help visualize what kind of connection types are being used to browse the web in Nigeria. From the charts, it’s clear that a good percentage of users browse the web using 4G speeds. On the flip side, over 50% of 170 million mobile devices use slower speeds.

While on a call with a colleague from Nigeria, we did a quick rendering test. He claimed to be using one of the best Internet services, but chances are it was still what I’d consider slow speed. My LCP element had rendered approximately 8 seconds before his had loaded halfway. I can only imagine how many “8 seconds” are lost when the web can be built much faster, in some cases with yesterday’s technology.

SEO, Engineering & Product Conflict

Having worked in the Web Performance space since SEO got involved, I’ve seen the hesitation from Engineering when it comes to SEO initiatives. I have been in scenarios where Core Web Vitals was placed in the same category as vanity metrics. I’ve also seen folks who have been in the Web Performance space for a while get worked up when other folks make it seem like Web Performance is mostly Core Web Vitals. I’ve seen how this dynamic makes it hard for Web performance tasks to get completed or implemented correctly. With all that conflict and friction in place, the actual web user is left with a poor web experience.

Querying similar data for the USA, focusing on the top 10000 websites, to compare the number of uses with 3G and slower speeds.

The amount of 3G users reduces by almost half. This should not be interpreted as over 70% of users do not experience 3G connection speed. All it takes to switch from a 4G to 3G connection in some cases is a trip to the basement. We cannot continue to build for faster speeds. We also cannot afford to only test on faster connections.

Hindrances

• If SEO is the blocker, talk ranking factor.
• If the blocker is a business/product, talk about the potential monetary value left on the table. (More ad revenue, more pageviews, higher avg time on page, high conversion rate. If you need to convince people, here’s a recent web.dev article showing how improvements to Performance had business impact.)
• If the blocker is with Engineering *cough cough* remind them about the engineering creed.

Web Performance tasks impact users in many ways. From getting to you through search to the retention of users and overall user experience. If you are still skeptical about Core Web Vitals and Web Performance tasks, remember the “mobile-first” initiative and how the web got better for users. We need to have a performance-first mindset going forward.

Let’s help make the web faster.