It's shocking how angry this article makes me. In an era of scummy ad-tech, decreasing search result quality, and explosion of crappy web content, I'd be embarrassed to publish this.
"I'm part of the reason you can't find what you're looking for but I'm getting richer for it, and for now so can you, until the whole of the internet is garbage meant for robots with no value to humans."
Not mine either, but I have noticed on HN that while there's plenty of deserved animus for the ills of adtech, there is little when it comes to the topic of generative AI.
We know that the lowest-hanging fruit use case for gAI will be to pollute the web further with machine-written content and clip art. It isn't to improve anyone's search results, it's to make somebody's bottom line go up, at the expense of the rest of us.
I wish more people would be upset about that. I mean we can already see what's happened to the web via SEO gaming. An even worse fate is coming once gAI really takes off.
"what's happened to the web via SEO gaming."
- The way I've been seeing it differs slightly...
Google's reactions to SEO over the years (and other reactions and actions) - has led to downranking so many good sites and props up sites that fit in with certain agendas.
Once more AI takes off, I expect many more niche search and find portals to be accessed on a regular basis which will bypass google's deranking and censoring, and also bypass any google SEO problems that people think exist.
The thing about SEO is, it’s only actually Google Search Optimization. Since Google has such a strong grip on search, the two are currently tightly coupled.
However they don’t need to be. If someone created a niche search engine it would immediately invalidate all of the Google-optimized trash out there. If it solves a specific use case people will probably pay good money for it.
In an honest form it’s “don’t confuse the crawler.”
In its actual form it makes the web worse. Even if you create a strategy that works Google will eventually shut it down.
It’s an adversarial relationship. The search engines want you to buy ads, not win shelf space.
The organizations that can pull off SEO still do it more to suppress competition. It’s not a low cost way to find customers anymore. It’s keeping people from figuring out you can file your taxes for free.
Legitimate SEO is still alive. Gaming SEO is now a losing proposition. I've told stakeholders for at least the last 5 years that writing genuine and relevant content is 90% of the job. Technical SEO is basically a task to not screw up and offend the crawlers. Web vitals is an interesting new twist but it's a better game than AMP was.
This is my impression / approach, too, and I’m wondering if someone here can bring a solid argument for a better option so I can have my team correct course or reallocate resources.
Ultimately I feel well-written and structured content is a self-solving seo scenario.
I think the game of "let's write a bunch of stuff so we get free views from Google" is also kinda dead. Your content has to be valuable and it takes a significant investment to do that. A good example is Hubspot which has dozens of articles on how to run digital operations for small businesses that are well-written, well-researched and contain some actual insights and even talks about products besides their own. If you don't have the chops to do that kind of thing, you're just better off paying for ad space.
Web vitals have very little impact on SEO though. Some of the top ranked sites have awful scores. Google has never said how much impact it has not I'd bet close to 5%.
We've actually seen web vital bugs tank our rankings pretty quickly. And fixing them resolved the ranking almost as fast.
I should specify that we are a timely news site, so there's some different mechanics at play. We mostly rank for content posted within the last 24-48 hours. Google has a slightly different set of rules for that kind of content vs marketing or evergreen.
Our biggest competitor fails in every Core Vital, and their page takes 13 seconds to load. Ours loads in under 2 seconds. They rank higher in every market. They run national ads and have tons of backlinks. Core Vitals aren't that important when compared to other things. If you're in a less marketable industry, your CLS being .01 better than your competitors isn't going to do anything. I can make my core vitals score absolutely perfect, but it could hurt user experience along the way. It's not a perfect metric by any means. Just something people bring up during sales pitches at this point.
It’s funny but websites actually compete with Google, since Google wants to keep users in the search engine and not clicking on a web search link, Google will summarize the site’s content to the search query such that the user will not clink on the link to leave the Google search engine
If I'm not mistake google doesn't get paid for ads unless you click on them. In other words google doesn't charge Ads customers for impressions, only clicks.
I was discussing regular web links. I think Adsense is still PPC, which is a bad deal considering it may not result in a conversion. I’ve only used it when they send me $100 free credits. It tilted in no sales. I actually sold a book in my website last through MySpace, so I’m probably not an expert on any of this.
I talked with somebody at Zapier a few years ago about freelancing for their blog. They said they’d pay me $100 for a first post and after 2 or 3 rounds of feedback from the editor I just stopped replying and lost interest. In hindsight I wish I had just asked for the $100 to see if they’d pay out for something that they thought (in my estimation) sucked.
Back in the dark ages (happy days), SEO was actually a real thing. You would use the HTML meta description, keywords, etc, to set up things which would actually be used by search engines to categorise and rank webpages.
Then gooble came along and decided to rank search results based on:
- The number of links back to something, which depended on popularity and longevity, NOT relevance or quality.
- Who pays gooble the most. Primarily who pays gooble the most.
Gooble completely did away with actual SEO. So now all these businesses promising to do SEO for other businesses are mostly just throwing links and blurbs on every social media platform (or Gooble Sheets etc), and maybe buying a page rank boost from gooble directly.
Merely adding keywords in page content is not SEO. Not the old (real) way, and not the gooble way.
I don’t think meta description or keywords would work anymore. Their problems were the original reason PageRank was developed in the first place. There will always be a very long tail of documents that match a query, so you need some outside information to rank their relevance and value. In fact, spammy keyword-stuffed articles are likely to be better matches since authentic content probably won’t be trying to game the ranking algorithm.
Unfortunately now that Google so thoroughly dominates search, their perverse incentive to rank garbage ad-saturated listicles highly is not mediated by user behavior (to leave). This warning was also noted in the original PageRank paper.
You could just as easily tell it as meta description etc. incentivising keyword-stuffing, and PageRank/'gooble' coming along and more reliably determining quality by a form of peer review.
