Co-Founder and Managing Director at https://xitter.cf/t.co/wFd0PeBwjl, an NYC-based Search Marketing agency. Former VP Growth/1st employee @Investingcom.
NYC
Joined February 2009
- Tweets5.2K
- Following1.1K
- Followers2K
- Likes10.9K
Pinned Tweet
My latest post: 9 Less Common, Yet Super Powerful Competitive Intelligence Tactics for SEO.
Mentioning @ahrefs, @DeepCrawl, @semrush, @SimilarWeb.
And no gap analysis, I promise.
link.medium.com/2LpDBSWgA1
If you are building decks today in any way other than with Claude, you are wasting your time.
Claude's latest ability to generate slides from scratch or make fixes has been incredible over the past few weeks.
About 400 resumes for an SEO role we published 2 weeks ago. Not sure if it's a bad sign for where things stand in the industry, or if there are simply way too many SEOs at this point.
400! We are a 9-person agency, not exactly a VP of SEO role at OpenAI, Google, or Zillow.
Most of today's AI search manipulations won't hold up. Once you get the actual mental model behind AI recommendations, this becomes pretty obvious.
Here's what I mean.
Before LLMs, if anyone asked you to find the best payroll software, you knew the drill:
Google it, build a shortlist, check listicles so you're not missing players.
Go to every company's site. Read who they are, what they offer, how they position, and compare pricing.
Then you go through testimonials, reviews, YouTube, Reddit, Quora, LinkedIn, Facebook groups, Trustpilot, whatever's relevant.
A day of research. Then you walk in with:
𝗛𝗲𝗿𝗲 𝗮𝗿𝗲 𝗺𝘆 𝘁𝗼𝗽 𝟯. Here's why the other 9 don't make the cut. Pricing, pros/cons, who to demo first.
This is exactly what AI does when it recommends a product. Same research and due diligence, but it comes back with an answer within 10-30 seconds instead of a day.
That's why we trust it. It's addictive, and it always has an answer.
Under the "GEO" hood, we have 3 main pillars:
— Tech
— Entity & site content (your own site being clear)
— Brand perception (reputation)
Many "GEO" services today focus on #3, with 2010-style manipulation. Pay to get listed. Pay to get recommended. Right now, it still works. But it has a shelf life.
LLMs move very fast, and they're getting better at spotting paid placements vs. authentic recommendations. Whether that gap closes in 6, 12, or 24 months, it closes eventually.
What won't expire: being a 𝗠𝗨𝗖𝗛 better product and making sure the internet actually knows it. Being what people really do recommend, everywhere someone asks.
There are two variables in this game that matter: being great, and being loud about it everywhere, including all places that matter.
A good campaign reflects that. It spots the gaps, makes sure the site is understood, and works toward the goal. Buying temporary placements isn't it.
You'd have to really underestimate AI to think fake recommendations will continue to work in the long term.
Exactly.
Nothing else, really, protects against core updates or SERP changes as well as branded demand.
"this is the insight I’d most want every founder to take away.Your branded search ratio is one of the most powerful defensibility signals you can show a buyer."
Vi the indispensable @igalst >> flippa.com/blog/how-search-v…
How to build a 'client brain' per account 🧠
My first piece on @sengineland is up: searchengineland.com/client-…
You know why? Because nobody wants 'em.
No branded searches = no demand
No demand = no reason for Google to rank
Nobody notices besides the brand and SEOs
Branded Searches was the biggest ranking factor 10 years ago, too. But thanks to AI, everything is now on steroids.
It's pretty much over for aggregator marketplaces #SEO
featuredcustomers .com - a B2B customer reference and review platform
over time 📉📉 .......
Well.. Hreflang could work here, @googlesearchc
Google isn't lying when they say: "We don't care if you produce content with AI". But that’s not the full story:
Over the past 2 years, every core update targeted 1 thing: Cleaning the web of URLs that add 0 additional value.
It's all down to that ADDITIONAL 1st-party expertise.
Very informative slide via Google's Danny Sullivan explaining the difference between "Commodity" vs "Non-Commodity" content
Google prefers the later
IMO, lots of evidence this is spot-on, not where Google is going in the future, but where it already is now
via @ChouinardJC
It takes the eye 0.5 seconds to recognize that a site was generated with Claude.
Vibe coding is cool, but you can take the extra 3 minutes and 22 seconds to make it a little bit unique.
Yes, and get rid of the icons.
Inspired by Roomba.
BREAKING: The FCC today officially granted @Tesla a waiver allowing it to use Ultra-Wideband (UWB) radio technology for its wireless EV charging system that will be used to wirelessly charge the Cybercab.
Normally, this kind of radio must be handheld and can’t be installed outdoors on fixed equipment. Tesla’s charging pad is fixed and could be outdoors, so they needed an exception.
The FCC said yes because:
• The signal is very low power
• It only turns on briefly while parking
• It works at very short range
• It won’t interfere with other systems
More information from the filing: "The Tesla positioning system is an impulse UWB radio system that enables peer-to-peer communications between a UWB transceiver installed on an electric vehicle (EV) and a second UWB transceiver installed on a ground-level pad — which could be located outdoors — to achieve optimal positioning for the EV to charge wirelessly.
Prior to the UWB operation, the vehicular system uses Bluetooth technology for the vehicle to discover the location of the ground pad and engage in data exchange activities (which is not subject to the waiver).
When the vehicle approaches the ground pad, the UWB transceivers will operate to track the position of the vehicle to determine when the optimal position has been achieved over the pad before enabling wireless power charging."
In its waiver request, Tesla states that the UWB signals occur only briefly when the vehicle approaches the ground pad; and mostly at ground level between the vehicle and the pad, and that the UWB signals are then significantly attenuated by the body of the vehicle positioned over the pad.
I don't know what happened to Twitter lately, but here's what I see:
1. 99% of posts I see are by accounts I don't follow.
2. 40% of posts are complete crap.
3. The other 60% is complete AI crap.
I know it's best practice and that everyone's doing it, but I just hate it. It wasn't cool 15 years ago, and it's not cool today.
Especially now that people literally miss what's an ad and what's not.
So GA4 isn't going to do anything by itself against all of that Direct, bot traffic from China, right?
At the pace things are going with LLM referrals vs. visibility, direct traffic will only continue to grow and at some point attribution will become a joke from the past.
Hi @asana, can my team please get back to work?
Asana's down for 8 hours now.
status.asana.com/
