Danny Sullivan presented at brightonSEO in San Diego, where he walked through the latest updates in Google Search. He gave insights into the search quality ranking system and addressed the audience's questions regarding ranking in the SERPs.
He started the presentation by saying,
“We’re not perfect. We make mistakes at Google. We don’t get it right. We would not do updates for assistance if they were perfect.”
He goes on to say that while nothing is perfect, Google aspires to get better at showing and ranking the most helpful content, which segues into the Helpful Content System.
As we know, the Helpful Content Update this last October shook up the SEO world. However, Danny says SEOs shouldn’t have been surprised, as the new system was built on the 2011 guidelines. He reiterates that any “new” guidance is built upon guidance that is already in place.
So why are people confused?
Danny says it’s because the guidance “don’t do things for Google” is counterintuitive to people who are trying to be successful with Google.
When looking at the Google guidelines, we need to interpret them as how people view our content, not the search engine. Additionally, there is no one thing that Google looks at to rank your content. Instead, Google looks at your behavior as a whole. So, if someone is recommending “the one thing you need to do to rank,” it should raise red flags.
He then goes on to define what helpful content is.
“Our systems are trying to reward what people like. If you chase the algorithm, then you’re behind. You should be chasing people.”
Danny explains that there are search and content cycles. When unhelpful content rises in the SERPs, the search engine will try to fix it. We are in one of those cycles right now. With this constant evolution, SEOs and search engines “have to keep evolving systems to improve satisfaction.”
Q+A With Danny Sullivan
How does Google view AI content?
“We focus on the quality of content, not who produced it. Use AI to provide people with unique, satisfying information.”
Is the web content accessibility guidelines a ranking factor?
“No, but it’s a nice thing to do for people. And if you’re doing nice things for people, you may align with our signals. You may also align with some of the signals we use for good page experience.”
How does SGE benefit content producers to create the best content?
“SGE is an experiment to figure out one way generative AI might help with search. SGE, just like search, shows links — and this is because people want more information that links provide.”
LMM can sometimes generate false information. What steps does Google take to reduce hallucination in SGE?
“By constraining SGE to specific tasks including corroboration, we're able to significantly mitigate some of the known limitations of LMMs like hallucination or inaccuracies. But it's also not perfect and has known limitations.”
SGE currently has an end date. What does that mean?
“Search Labs experiments are..experiments! Experiments can be extended or could end early. End date does not mean an experiment will launch into search right away or at all in its current form. We're still experimenting with SGE and it's changing rapidly.”
Can I block SGE from using my content?
“AI is built into Google search in various ways including SGE which is why you'd have to block Google search. But also as being a part of Google search SGE does follow non-snippet and max-snippet.”
If Google matches content to the terms that are not part of the search query, what shows up in Google Search Console reports?
Search console shows the query as entered. This makes sense because even if we go beyond the terms it's still the query that was entered.”
Why does Google rank terrible content by big brands over better content from smaller brands?
“We’re not perfect. Sometimes, we also might do the opposite. We understand the ecosystem concerns that big sites shouldn’t be seen as ‘getting away’ with things. Our goal is to ensure we’re equitable in ranking — and we’re still working on that.”