Page 2 Podcast | SEO Podcast

Page 2 Podcast | SEO Podcast

@page2pod.com

The Page 2 Podcast is an interview show about the people doing SEO and AEO work: the experiments they've run, what the data showed, and how they got here. Hosted by @jonclark.co and Joe DeVita

37 followers221 following57 posts

Angela Skane on the Page 2 Podcast: she would rather optimize the content she already has than publish 20 more articles. 100 pages sitting on page eight of Google do not make a site stronger, they just make it heavier. Has a publishing quota ever worked for you?

The hard truth for clients: traffic isn't coming back. SEO was rankings and traffic; AI search is visibility and downstream conversions. People research in the LLM, exit, then run a branded search. How are you handling attribution in answer engine optimization?

New episode is out now: Marcus Tober on why rank tracking is harder than ever. num=100 is gone, the referrer is obfuscated, and he believes AI agents send 1,000x more requests than scraping. #RankTracking #SEOTools

Two AI tools worth trying daily: Fathom for call notes so you stay present, and Whisper for voice-to-text everywhere — Slack, terminal, even 'voicing' code instead of typing it. Voice AI is a genuine speed unlock. What's the one AI tool you can't work without?

New episode of Page 2 Podcast is live with Greg Gifford. He breaks down what actually proves local relevance to Google across sales, service, and parts departments, and why category selection matters more than most dealers think. What's the local SEO myth you hear most often?

Sophie Brannon's whole culture-fit screen is one question: what are you most passionate about? A work answer, or anything obviously tailored to what she wants to hear, is an immediate no. Some of her best hires said gaming. She notes it works on dates too.

Vertical video isn't ranking in AI Overviews yet. Cindy Krum, CEO of MobileMoxie, explains on @page2pod.com that Google tends to test new formats like images and video shorts in their own tab before folding them into search. Integration into AI Overviews looks like a matter of time.

Distribution isn't one-size-fits-all: once on LinkedIn/Facebook, three times on X by time zone (a reader poll confirmed people want it), with different headlines and images. YouTube gets A/B-tested thumbnails and titles. How do you handle reposting the same piece?

A deliberately human-first choice: we don't automate call analysis with AI. When you automate something, you stop paying attention to it. A person listening to calls and taking notes is what sets the work apart. Where are you keeping humans in the loop on purpose?

View full thread

llms.txt. He's never seen one increase visibility in an AI answer. Its real use is pointing actual agents at your API, MCP and docs. What does work is unglamorous: describing your brand the same way on your site, LinkedIn, G2, press releases. LLMs want consensus.

Markdown copies of your HTML pages. He ran it across hundreds of thousands of chats and found zero cases of an LLM citing the .md version over the HTML page. Cost of shipping it anyway: - duplicate content - twice the crawl budget a diluted internal link graph

Structured data as a fact pipeline. In a published test, when the page text and the schema disagreed, the text won. When the fact lived only in schema, ChatGPT answered ~20% of the time and Claude 0%. Schema still helps. Just not as a place to hide information.

An early SEO-testing finding for LLMs: the winning tests are about giving more information. LLMs don't get bored with long specs the way humans do — manufacturer data, product detail, and exposed info all help. It doesn't have to be UGC. What's moving the needle in your LLM tests?

AI referral traffic is sitting at 1.08%. Pat says every study keeps confirming the same thing: this was never a traffic play. What clients actually ask about is brand sentiment and visibility. He thinks the number only moves once agentic ecommerce puts clickable product cards in the conversation.

A technical SEO trap for news and evergreen content: Google often pulls the wrong date — from structured data, on-page, or CMS sections — sometimes off by years, and it kills clicks. Show one clear date, not multiple. Have you been burned by Google surfacing the wrong date?

Page 2 Podcast | SEO Podcast reposted

Super excited I got to chat with @jonclark.co and Joe DeVita on the @page2podcast.bsky.social!! We've all seen people scramble to explain why performance is down and sometimes just "accept" good results, but if you can’t explain the win, you can’t replicate it. youtu.be/YrE21LqTU2o

youtu.beDirect Traffic Is Your Missing AI Metric 📈 - Derek Hobson on Why You Should Report SEO + DirectYouTube video by The Page 2 Podcast: An SEO Podcast

A nuanced read on the Google clicks debate: Search Console shows clicks down, Google says up — and both can be right. At trillion-URL scale, their vetted statement reflects global data; your individual site can still contradict it. Different vantage points. Fair framing, or a cop-out?

Fascinating indexing strategy: a self-learning model, built over two years, decides which URLs are indexable using interaction and demand signals - and if a page clears the bar for the site, it usually clears Google's index too. Would you let a model gate your indexation?

Ten years before it was consensus: a team cut 80% of a client's content and rankings improved across everything that remained. That's the origin of 'quality is a site-wide metric.' Google's court docs later confirmed site authority exists. Have you seen pruning lift a whole site?

A framing worth sitting with: AI literacy is becoming like internet literacy: a skill you're economically locked out without. The way to build the muscle is to treat it like a person and pressure-test it constantly. How are you building AI literacy on your team?

Pat Reinhart's AI literacy interview is just: here's a task, build an agent with a prompt that completes it. Takes minutes to see who has hands-on reps. His prediction is that this ends up as baseline as "proficient in Excel" was twenty years ago.

Angela Skane made a point on the Page 2 Podcast that stuck: one idea can carry a blog post, a YouTube video, shorts, an SME LinkedIn article and a guest post, with the brand narrative intact. Which format do you always skip because nobody owns it? New season 6 episode out now!

On content in regulated industries like legal and health: center their fears, treat the constraints as a box, and shape your strategy within it. The output won't be exactly what you envisioned, and that's okay if it still serves the user. How do you work with compliance?

A Reddit playbook for brands: you need three accounts. Your own subreddit (claim it before someone else does), a brand account for the nice mentions, and a personified human account for complaints and questions. People do business with people. How's your Reddit presence structured?

The website in a zero-click world: relatively less important, absolutely still essential. Don't cut investment in your own site but do invest in AI overviews, what ChatGPT says about you, and marketplaces. Zero-click marketing is the job now. How are you adapting?

A team-building take for AI search: don't hire AEO/GEO "experts" with no SEO foundation. Everyone on the SEO team already does AEO work - you evolve their skills. The real shift is closer work with social and more in-house digital PR. How is your team adapting? Full episode with @lilyray.nyc

Googlebot renders a page at a fixed viewport, then checks the page height once during rendering. Infinite scroll and lazy loading pages that need multiple scrolls to reveal content mostly lose everything past that first check. If you work in technical SEO or JavaScript rendering, worth a listen!

Agencies stuck at 20-40K/month usually have three problems at once: undercharging, over-delivering with no clear SOWs, and over-compensating team or self. Fix all three and gross + net profit jump. Which of the three is hardest for owners to face? Full episode with @johnfdoherty.com