1) https://www.webpagetest.org for analyzing loading times
2) a3 lazy load for lazy-loading images: https://wordpress.org/plugins/a3-lazy-load/
3) autoptimize for combining multiple CSS files: https://wordpress.org/plugins/autoptimize/
More description of what we did at my write-up here: https://hackernoon.com/dont-brake-for-fonts-3-web-performanc...
In my experience the biggest issue with with slow load times are people uploading large unoptimized images. Things like multiple css/js files, unnessary plugins and framework are nothing compared to people using uncompressed large images. Use a plugin, ImageOptim or cloudinary.
It loads the page and the text first so you can start reading. The images and scripts are delayed and cached. Comments are last. Pictures and thumbnails have various resolutions to improve load time.
This change is more focused on mobile.
[0] https://webmasters.googleblog.com/2010/04/using-site-speed-i...
> The “Speed Update,” as we’re calling it, will only affect pages that deliver the slowest experience to users and will only affect a small percentage of queries.
The desktop ranking signal had almost no effect; slow pages reliably turned up in search results. Maybe some pages that took 15s to paint were slightly diminished in results, if that.
If the mobile ranking signal is as weak as the desktop ranking signal, (and they seem to be indicating it's going to be a very small effect,) I would expect this change to have no measurable effect at all on the median performance of pages loaded from Google search results.
Using a CDN is one way to speed up your site, but there are lots of other important steps like reducing dependencies and compressing your resources.
(Source: I used to work on mod_pagespeed at Google, now I work on other things there.)
Kind of like how amazon shows you the search results by "Relevance", "Price", "Popularity", etc. Maybe google can do it too and let us sort by "Relevance", "Speed", "By Geography", etc.
Concerns over Google-the-company aside, Google-the-search-platfrom really sucks IMO because it assumes it knows what I want, rather than just asking me. That mentality was Steve Jobs' worst legacy and I hate how it's wound up everywhere.
* Prioritize fast-loading pages?
* Prioritize mobile pages?
* Prioritize by pagesize?
* Prioritize your language? Prioritize local results first? Prioritize news sites? Prioritize blogging sites? Prioritize comment platforms? Prioritize "safe-for-work" websites? etc...
There's an almost unlimited number of questions that you could be asked, and the reality is that nobody wants to answer any of them.
At some point, they need to make some decisions for you, otherwise you are tasked with coming up with your own search algorithm yourself! The question becomes which decisions should they make for you, and which should they let you make.
Currently the answer looks like they are making most of them for you, while allowing you to make decisions on whether you want "web, videos, images, books, news, etc.." and what date range you are searching for.
I think that's a good choice for them, because at the end of the day if you want more control or choice in your searches, you can always use another search engine.
This isn't also unique to Google or their products. I've struggled with this in my own programs and products, even open source ones. An endless list of configuration options means that nobody will ever set them all correctly, and it increases the amount of testing you need to do almost exponentially. Choosing good defaults, or in many cases making the decision outright for your users is the best possible case, because there just isn't any other way which scales out to the entire piece of software in most cases.
Now I am old and been using search engines for a long time and remember Veronica and Jughead coming online and do think I have learned through the years how to word a search to optimize getting the results.
