5 ms·
> But AMP is not necessary to make a fast page. That's true, but almost no non-AMP news sites aren't total garbage (and some of them have even manage to make t
by s17n 6y ago
> But AMP is not necessary to make a fast page.
That's true, but almost no non-AMP news sites aren't total garbage (and some of them have even manage to make their AMP sites terrible). Without Google taking an aggressive approach, things will get worse, not better.
- ballenf 6y agoI don't think anyone (here) would complain if Google explicitly used Lighthouse scores in ranking algorithms. Or any similar agnostic user experience ranking.
- s17n 6y ago"We need to make an AMP site to get into this carousel" is a much easier sell (to executives) than "we need to rebuild our site according to some not completely specified principles to hopefully get a higher search rank".
- stephenr 6y agoIt really can’t be hard to set a metric. Time limit (load time and render time), and to a lesser degree, weight limit (ie total resources size). This is hardly a new concept.
- josephg 6y agoYeah; and the spectre of page weight might be a good motivating factor. "Each useless tracker we add lowers our search rank. Are we sure we need to add this one?"
- IanCal 6y agoLoad time in what browser? Over what connection? When is it loaded - all pictures or just the text? How much text? Does all the JavaScript have to have loaded? What about the calls that makes, do they need to complete? Is it better to have one more second before first render or one more megabyte of data? I don't think it's a simple problem.
- markdown 6y agoIt's a multibillion dollar company with thousands of the best tech minds in the world in their employ. With the massive pool of talent they have, this stuff isn't exactly rocket surgery.
- IanCal 6y agoThere is no single clear metric, that's the issue. It's a lazy approach to just say that problems must be simple because they're a big company.
- stephenr 6y agoFine, if you can't pick a number out of your ass and say "sites must load in less than 1.5 seconds from <simulated average connection> to appear in our special box" just use it as a ranking value. Fill the special box with relevant results to the search, and use load time for the result as a weighting. So if your content is ridiculously on-topic but a little slower to load, you'll still probably be in the special little box, but a slightly less specific site thats super fast to load might be before your site. As has been mentioned: google has used page load times as a metric for a long time. The difference is they didn't add a "special little box" to incentivise sites with otherwise shit decision making skills to do the right thing. But sure. Tell me how forcing clients to download a bunch of javascript, and introducing forced 8 second blank pages for anyone who dares to not load said javascript is all about making pages faster. Give me a fucking break. If you want to live in the fucking google sphere, that's your choice, but don't pretend that their motives are anything but hostile to the very concept of the open web.
- IanCal 6y agoAn unnecessarily aggressive response, containing a bunch of things I've absolutely not said.
- s17n 6y agoGoogle has already done this, it hasn't had anywhere near the impact that the AMP carousel has.
- blackoil 6y agoSEO optimization is a large industry. n% boost in organic traffic is very measureable.
- sdfhbdf 6y agoThey will in next year - it's called Core Web Vitals [0]. [0]: https://developers.google.com/search/blog/2020/05/evaluating-page-experience https://developers.google.com/search/blog/2020/05/evaluating...
- underwater 6y agoWhen I worked as a publisher AMP pages were slower to render than our web pages, though faster with the head start Google gives them. Google could introduce rewards for fast web pages too, but chose not to.
- s17n 6y agoGoogle already does reward faster pages and publishes tools for benchmarking yourself but it hasn't had anywhere near the impact that the AMP carousel did.
- MaxBarraclough 6y agoDoesn't this tell us that their penalty on slow pages just wasn't harsh enough? If publishing slow, bloated pages would reliably come at the cost of not being featured in the first page of Google results, sites would soon change their ways.
- s17n 6y agoMaybe, maybe not. Rewriting your existing site is much riskier for an organization than adding a new site which gets its traffic from a new source, while your existing site continues to chug along as normal (albeit with gradually declining search traffic from Google as the AMP box takes more and more of the share).
- MaxBarraclough 6y agoI don't agree with your uncertainty here. If Google imposed a seriously strong penalty against bloated websites, people would promptly stop making new bloated websites, and would invest in lessening the bloat of their existing sites. I can't think of a more certain spur to action than Google imposing a threat of going undiscovered. It's perfectly normal for web companies to flout the GDPR, say, as it's laughably unenforced, but Google rankings are no joke. edit: a few tweaks
- acdha 6y agoWashingtonPost.com consistently renders faster than AMP for me. The main thing is that they have a solid engineering culture and thus don’t have a ton of JavaScript in the critical path to rendering so you don’t get the AMP behavior of showing a blank page for many seconds when anything goes wrong loading Google’s multiple large JavaScript files. If you use a mobile device this happens surprisingly frequently - it was one of the factors motivating me to switch to DDG since Google favoring AMP meant I’d hit blank pages multiple times a day.