Oldest MYTHS in SEO revolve around this subject.
How many words should a post be?
As many as it takes to cover the topic and meet site goals.
SEO Industry obsession w/word count is just cookie cutter/checklist SEO taken to an absurd extreme! Pure nonsense based on a COINCIDENCE!
Agreed, but if you did it immediately you were building a separate mobile site and that eventually was a waste because the tech improved to where responsive design worked.
What many haven't considered is that many users enjoy the browsing experience, don't trust AI cuz of hallucinations & FOMO
The smart move is don't change anything pre-migration cuz if there's any issues the fixes are easier to spot. Also once it's migrated it's easier to access the outcome from one change rather than figuring out the cause amongst multiple changes
Agreed SEOs haven't widely adopted product feeds as a strategy, however, I'm skeptical of AI shopping as the future of E-commerce SEO. When I see ACTUAL USERS touting it my skepticism will lower.
IMO, people you see making these statements HAVE SKIN IN THE GAME and are biased!
SEO is always about subtleties! It's knowing where you cross the line from optimization and maximization to GREED & SPAM! Executives/owners seldom see the line... that's why they should hire SEOs who do!
I agree w/most of this as it applies to grounding etc. I agree that AI is much more interested in verification and provenance of facts and comparisons, however, Schema and structured data of any kind is also good for entering the "conversation"/training LLM it's easier to digest same as markdown is
Ummm Bill Slawski has never had one... so I'd say that is about right... only journalist/publisher who should maybe be there is Danny and more because of his role in starting the major conferences and his work at SEMPO! Same for Brett Tabke for PubCon
IMO, users who would replace traditional search and BROWSING TASKS w/AI feel much differently about AI then those who use it for work. TRUST (or lack thereof) amongst these users is a bigger factor than FOMO.
It will be years before AI agents are doing these users purchasing for them!
Not a comprehensive study but interesting insights into when search is triggered, content strategy for multiple AI platforms, how OpenAI had baked-in brands
zkami.substack.com/p/i-ran-8-in...
Pedro has written on this topic a few times. It says a lot about SEO industry when Prompt Tracking becomes a KPI. One of the reasons I stopped taking clients was there are no SUCCESS KPIs that indicate the TRUE VALUE of the work SEOs do! Prompt tracking is a joke!
theinference.io/p/half-your-...
Not new! This one has SUPPOSEDLY been on the books for decades. First time they imposed it we heard a few complaints...
IMO, it is no where near as bad as it was when they first announced they were going after this nonsense.
IMO, AI task completion is all hype by folks who benefit from the hype. FOMO is a real motivator for browsing not to mention trusting an agent to find cheapest price or "best" product when it is clearly manipulated by false claims of "best" on the many sites abusing listicles!
Mentions and citations are not determined algorithmically they are probabilistic calculations. To call this "ranking" is a joke it's more of a guess with no rhyme or reason not to mention a high % of citations ARE HALLUCINATIONS!
Not surprised if most of the losers were either victims of HCU or would be now. Every one of the characteristics you listed were on my HCU list of do/don't list. Topic drift or topic relatedness is measured mathematically using embeddings so what seems logically relevant to humans often isn't!
Anyone who thinks this will be successful... is going to be greatly disappointed! They provide no analytics so... it's a suckers bet if I've ever seen one! You're basically buying a PIG IN A POKE!
Agreed but in cases where JS & CSS are in security sensitive folders still block folders but ALLOW JS and CSS and test it because sometimes the order of robots commands can affect the ability for those files to be indexed.
I've seen lots of cases WP admin folders exposed for JS and CSS indexation.
Well of course not! Prompt tracking is useless as tits on a boar! Anyone paying attention should have realized these aren't based on algos but random/probabilistic calculations! Prompt tracking as a success KPI is like playing Career Russian Roulette w/5 bullets in the chamber!
Every person doing GEO needs to read this!
theinference.io/p/geeeo-was-...
VCs have no problem saying anything to sell what I believe are useless tools. Prompt tracking is even less useful than ranking reports. They provide ZERO actionable data and provide nothing to address the issues they uncover!
Barry is one of my faves and one of the few I pay attention to when he speaks. The seduction he talks about here is real because it's easy to lose your way when you earn JUST BY GETTING CLICKS.
Golden Rule: Google giveth for quality and it taketh when quality fades
If you looked at traditional search the same way inaccuracy on the sites in the results would be about the same. Nothing on the web should be considered accurate w/o verification. When I review AI anything not cited is treated as suspect until...
"If you’re not in the conversation, you’re not in the answer."
Definitely getting no mentions or citations!
.co is the TLD for Columbia (sure you knew that) so if the AI hasn't figured out that several sites use it as meaning "company" I can see how it thinks it suspicious. IMO, Google has for years treated some TLDs as junk, spam or suspicious.
An advertorial the publisher should be ashamed to publish BS like this! ChatGPT pound the FCK out of servers because they don't have an index which according to the author is a good thing!! Take everything and give next to nothing back!
Nonsense! As a longtime one man operation almost everything AI is doing I could program decades ago! AI just puts that ability in the hands of people who do not have the programming skills however, they don't really have the KNOWLEDGE to manage and plan these websites so it doesn't implode later!
SEO Gurus are generally false Prophets!
Reminds me of when Google used similar tactics to read bookmarks which are definitely in a sandbox and out of bounds and As far as we know stopped soon after. Extensions are in a sandbox as well, however, it isn't as strict and it's more about lack of ethics than invasion of privacy.
Prompt tracking is AI equivalent of a ranking report. Provides no actionable data beyond indicating if you have citations and mentions, however, these both seem random w/frequent changes
Better to determine if prompt is:
grounded
get RAG searches
you have included tokens?
These are actionable
IMO, many you can't really blame the ump as several I've seen were less than 1/10 of an inch. Only seen a couple that I'd say were a bad call by ump
Yes and no one it seems thought it through far enough to realize that results for Bing and Google often overlap so just seeing url is in Google doesn't necessarily mean the search wasn't on Bing and url is in both results
Not to mention the grocery idea the NDP are suggesting is just what partially got Mamdani elected in NYC! Just another bunch of idiots like the PCs who take the worst ideas of the US Republicans and the Trudeau Liberals who tried to shove Democrat WOKE nonsense on us.