/seo/ - SEO Techniques

>>2145
openai isn't building their own index from scratch, they're just scraping everything and it's likely just some massive lag in how they process 'newly deindexed' signals.

>>2145 had a similar nightmare with a local law firm where 'google search console' showed zero drops but all our referral traffic from ai bots was just garbage pure nonsense.

>>2142
$148 million is such a ridiculous number to throw around. even the kerning in that title looks suspicious.

>>2151
the kerning is definitely a red flag but i'm more hung up on that huge valuation claim because it reminds me of this one agency i worked with last year that promised us massive organic growth by basically just automating thin content. we spent months following their "proprietary" framework only to realize they were just spamming keywords and hoping the algo wouldn't catch them immediately. seeing them push a whole book on generative engine optimization feels like the same pattern of trying to manufacture authority where there isn't any real substance. it's hard to tell if people are actually learning anything new or if we're all just being sold a shiny new way to do the exact same old tricks. have you noticed anyone else actually seeing organic traffic stability since these ai search updates started hitting?

>>2151
>even the kerning in that title looks suspicious
The spacing is fine, it's just a badly designed thumbnail for their ad.

>>2143
google's just trying to kill all traffic now. ai summaries are a joke and we're all just feeding the beast for nothing.
>adding more layers for ai discovery feels like extra too much work
it's not even about discovery anymore. it's about zero-click cannibalization. i'm seeing the same thing with my health clients. if you can't get a snippet, you don't exist. are you using any specific schema markup to try and force those clinician names into the snippets? im asking because structured data is getting ignored more often than not lately.
good luck with that dentist

>>2149
try doubling down on author schema for every single bio page, but it won't help if your content is too thin to even trigger a snippet. tbh i've seen medicalEntity markup work okay for some clinics though.

>>2138
>hard to see how i'd even track if we're actually being cited by perplexity or chatgpt without more tools

tracking isn't the hard part when you check your referral traffic. just look for the direct hits from those domains in search console and see if the bounce rate is garbage. if the traffic is coming through but users aren't clicking anything, then you've got an issue with the snippet content itself. it's probably way more about content relevance than finding some magic new dashboard. why would a plumber even care about perplexity citations when people just want a phone number?

>>2140
>if the traffic is coming through but users aren't clicking anything, then you've got an issue with the snippet content itself

checking bounce rate works for the bigger sites with high volume but it's pretty useless for a small local plumber because the sample size is way too low to be meaningful. i usually just use a simple brand mention check in search console to see if our site name is showing up in queries that trigger those ai overviews.

>>2144
>i usually just use a simple brand mention check in search console to see if our site name is showing up in.

brand mentions in gsc don't tell you anything about ai citations. search console only tracks clicks and impressions from the actual serp, not when some LLM scrapes your text to answer a query. ime i check for increased direct traffic alongside weirdly specific long-tail queries that shouldn't be ranking.
google is just rebranding the same old scrapers

>>2128
>i don't see how walking around looking at physical posters helps anyone when we could just watch a video
the poster sessions are basically just an excuse to network with people who actually know their stuff instead of sitting through another generic keynote.

>>2133
>the poster sessions are basically just an excuse to network with people who actually know their stuff
networking is literally what the cocktail hours are for, posters are just a waste of floor space lmao

>>2135
>networking is literally what the cocktail hours are for
the posters are for actually seeing the data without some speaker's slides failing mid-presentation. i'd rather see a clear breakdown of crawl budget changes on a board than listen to some guy drone on about his "ai strategy" during a drink. most of these sessions end up being nothing but pure marketing fluff anyway.