@RyanJones Probably one of the toughest things anybody has to go through. Take care!

@darth_na @brodieseo In most cases I've seen, it's not what either the site or the user would want. But if there's 1 thing I have confidence Google is good at, it's showing the right type of page to the majority of users. If they bubble up a category over a product, it's likely a better result. YMMV.

@darth_na @brodieseo A1: more often than they should or used to. I'm inclined to think they dropped the ball on paginated archives, and their heavy lean into ML now results in subpages now outranking the head page in certain cases. I've seen it much more. That said, canoniclisation not solution imo.

@wilreynolds Never mind AIO. The CTR for organic versus paid is staggering. Is this analysis excluding clicks to other Google serp features e.g. refinements etc.?

@dejanseo Do you use with Cursor or Windsurf for no-code, or just coding to their API?

@dejanseo V interesting. Nice work. I wonder how they are interpreting the response? It feels likely it's fed into another prompt? "The model receives clean text from Chrome". What does the function they use look like? Do you know the specific function call associated with extraction Dan?

Google News challenge with syndicated content. How many duplicates can they include for a single topic?

@Adoubleagent I have to say, I find your tool one of the more fascinating I've seen in recent times. Simple solutions are very often the most useful. Well done.

@ziodave @darth_na @dejanseo @cyberandy @timberners_lee @oralassila @jahendler Do you know of any research around this?

The presentation layer will become increasingly redundant. Interesting given yesterday's chat with @dejanseo and @cyberandy around SD and KG. Certainly if publishers/sellers get onboard, and standards are used/created, SD becomes very powerful indeed.

@rustybrick @LauraChiocciora @glenngabe @fabienr34 Worth noting it was already invoked in Europe (UK and IE sites previously hit) and this looks very like it's expanded to non-EN countries.

@cyberandy @marthavanberkel @vberkel @dejanseo Please share if this is recorded. Would like to catch it after if I can (scheduled my night time).

@cyberandy @dejanseo @timberners_lee @oralassila @jahendler All absolutely sensible. But coming back to the agent running on my phone and gathering data - how will it know which data to trust? Can we break out of Google's walled garden? Will the only agents be Google agents? Does SD help or hinder?

@cyberandy @dejanseo @timberners_lee @oralassila @jahendler Totally concur. I think there are overwhelming benefits for publishers and content owners to introducing KG and ML for internal processes. But now think of it from a spammer's POV. How will all these agents know who to trust? I think this will be a very serious challenge.

@cyberandy @dejanseo @timberners_lee @oralassila @jahendler This is where Search Engines like Google have a leg-up IMO - they have dealt with the spam and crap for many years, whereas these new agents are going to need to learn very quickly. The big challenge I see with agents consuming SD is the very vast swathes of shit online.

@dejanseo @cyberandy Andrea's opinions most welcome here. I've conversed several times with him on this in the past, and always enjoy his thoughts around this. I know he's building KGs and agents, but I wonder how much he'd trust 3rd party SD?

@dejanseo Always susceptible to rubbish-in, rubbish-out. It will make sense so much as the data is accurate, reliable and useful. I very much concur with data becoming the main currency, but I'm not so sure that publisher-driven SD will make the cut. Perhaps.

@dejanseo Small leap here to connect this to SD?

@RyanJones How about re-evaluating updated page without re-running the SERP analysis? Possible? How can a user know the updated "score" of their own page? Sorry for lots of questions - curious how you've built this.