Eli busts the belief that rising SEO traffic is inherently good. Unless you're a media company monetizing pageviews, more traffic that doesn't convert to leads or sales is worthless. He recommends deleting content on the non-converting side of a business and measuring SEO against real business metrics like MQLs, not rankings.

Rankings and raw traffic are the wrong success metric for most businesses
He once advised deleting all content on the non-monetized side of a two-sided marketplace
The myth that Google rewards you just for having traffic (or a blog) doesn't hold
Pick a conversion metric that matters: MQLs, dollar conversions, phone calls, or pageviews only if you're media

“if you are driving leads and that that that traffic number goes up you're not driving more leads from SEO you're getting worthless traffic it…”
Eli Schwartz · 51:30

#conversion#metrics#seo#traffic

Myth Buster1:07:30

The Biggest SEO Myth: Link Building Is the Secret Sauce

Eli calls link building SEO's biggest myth. The usual tactic, buying guest posts and links on low-authority sites nobody reads, doesn't help. The right approach is to build a brand people actually want to use and mention. Post-LLM, Google can read a mention as a signal even without a traditional HTML link.

Most link building means buying guest posts or links on low-authority sites nobody reads
It's ironic to call Google an all-knowing AI while assuming it falls for those cheap links
Build a great product that earns mentions; link building is brand building
Post-LLM, a mention (not just an HTML link) can count as a brand signal

“the way most people build links is they buy these guest posts or they pay for links on low Authority websites”
Eli Schwartz · 1:08:00

“link building is is brand building so it's thinking about how you're going to build a brand and have people mention to you”
Eli Schwartz · 1:27:30

#link-building#brand#seo#myths

Myth Buster1:37:00

Keyword Research Tools Are Wrong by 10x in Both Directions

Eli warns that keyword volume tools (Semrush, Ahrefs, Similarweb) are not sources of truth: for real client keywords he's seen them over- and under-estimate volume by 10x. Even Google can't share its own ad numbers for legal reasons, so every tool is guessing with proprietary algorithms. Use them for normalization and relative comparison, not as forecast inputs.

For the word 'WordPress', every tool's number was completely wrong versus Google Search Console
Tools have over- and under-estimated real keyword volume by roughly 10x
Google is barred from sharing its true ad-search numbers, so tools buy data and estimate
Use tools only for normalization and relative comparison, not exact forecasts: trust Search Console

“the number that every single keyword research tool had was completely wrong of what Google search console said for the word WordPress”
Eli Schwartz · 1:40:30

“some of them the tools were overestimating by 10 times sometimes they're underestimating by 10 times”
Eli Schwartz · 1:42:00

#keyword-research#forecasting#tools#seo

Hot Take· 3

Hot Take35:30

Most SaaS Companies Shouldn't Do SEO At All

Drawing on Mixpanel and Google Cloud, Eli argues that many SaaS products have no real SEO journey. If a tool requires company-wide integration, a big budget, and a committee decision, no one converts from a search result. Making people aware the problem exists is a brand challenge, not an SEO one.

Mixpanel's content ranked and got traffic but didn't convert: it's a company-wide, expensive, high-friction purchase
Google Cloud has only two competitors and buyers make committee decisions, so SEO makes little sense
SaaS solves a problem only once you know the problem exists, and creating that awareness is a brand job
SEO 'plateaus really quickly' for products with no genuine search journey

“getting people to know that problem exists is typically not an SEO challenge it's typically a brand challenge”
Eli Schwartz · 36:30

“no one is going to do a search and be like Oh Google's number one for this term I searched let me just go buy…”
Eli Schwartz · 38:00

#saas#seo#b2b#brand

Hot Take1:05:30

Showing Up in AI Overviews Is a Branding Win, Not a Content Win

Eli treats AI Overviews as a branding exercise. If your link is summarized, you likely already ranked beneath it, and Google may just be lifting your content for free. But if you show up named as a brand ('the top CRM tools'), that means your brand-building worked, which is the outcome you actually want.

Google's in-answer links may exist mainly for liability protection, not to send you useful clicks
Eli's LinkedIn survey found link clicks roughly 50/50, more than he expected
AI Overviews essentially duplicate the search results sitting right beneath them
Showing up as a named brand = good branding; showing up as a link = Google took your content

“I think of AI as or AI overviews as a branding exercise”
Eli Schwartz · 1:05:30

“if you show up as a brand you've done good branding you show up as a link it means Google stolen your content and people…”
Eli Schwartz · 1:07:00

#ai-overviews#brand#seo#google

Hot Take1:16:00

Why Google Won't Be Dethroned: Lessons From the 286-Page Verdict

After reading the entire 286-page antitrust verdict, Eli argues Google's dominance is nearly unassailable, not on quality alone but on distribution and habit. The verdict pegs Google at 98% of mobile searches, shows how default agreements (Apple, Chrome) crushed a better search engine (Neeva), and notes DuckDuckGo went from 1% to 2% share over a decade despite huge spend.

The verdict states Google has 98% of mobile searches
Neeva had a better engine but failed for lack of default distribution deals
Apple refused to take Bing at any price; a Mozilla-Yahoo switch saw users revert to Google
DuckDuckGo moved only from ~1% to 2% share over ~12 years despite ~$100M raised and heavy branding

“in the in the the verdict it says that Google has 98% of mobile searches”
Eli Schwartz · 1:17:30

“I don't think there's a chance that chat gbt or perplexity or Claude or any of these other llm startups”
Eli Schwartz · 1:19:00

#google#antitrust#distribution#market-share

Explainer· 4

Explainer07:30

Why Google Rushed AI Overviews: Wall Street, Not Users

Eli explains that ChatGPT's late-2022 arrival forced Google's hand, but the real pressure came from investors. If Wall Street decides Google is a has-been, its stock falls, hurting its ability to recruit and invest, so Google had to publicly prove it could do AI too. That urgency, not user demand, is why it launched Bard (badly) and then AI Overviews.

ChatGPT sparked a narrative that people no longer need Google or SEO at all
Eli's core stakeholder argument: Google answers to Wall Street more than to searchers
A falling stock price would hurt recruiting and Google's ability to fund its ambitious projects
Google's first response, Bard, demoed badly and its stock actually dropped before it recovered

“one of Google's primary stakeholders is really Wall Street so if Wall Street suddenly thinks that Google is a hasb been company and they're not…”
Eli Schwartz · 07:30

#google#ai-overviews#seo#strategy

Explainer09:00

The Three Problems Google Still Hasn't Solved With AI Answers

Eli lays out why Google agonized for a year over AI Overviews: monetization, liability, and plagiarism. If Google shows an AI answer it can't also show the ads that make up most of its revenue. And once Google itself becomes the publisher of an answer, it inherits legal exposure for bad advice and for lifting others' content.

Monetization: an AI answer crowds out the ads that drive the majority of Google's revenue, an issue Eli says is still unsolved
Liability: Google becomes the publisher, so a harmful generated answer (e.g. the 'jump off the Golden Gate Bridge' screenshots) is now on Google
Plagiarism: Google is extracting and summarizing content it may not have the right to
These concerns are why launch-fast Google took roughly a year to roll it out

“if you're going to show this AI answer on a search result then you also can't have ads you need to go all in on…”
Eli Schwartz · 09:00

#ai-overviews#google#monetization#liability

Explainer11:30

AI Overviews Don't Hide Your Content, They Make It Irrelevant

The common fear is that AI answers push your links further down the page. Eli reframes it: the real shift is that the top-of-funnel discovery step gets swallowed by the AI answer, so ranking #1 on a broad keyword becomes less relevant, not just less visible. The winner of the long-form content race no longer gets the first click.

The problem isn't that your content is harder to find, it's that it becomes less relevant
Broad top-of-funnel keywords are exactly what AI Overviews are best at answering
The discovery step moves to the LLM; the searcher only returns to Google once they know what they want (mid-funnel)
Whoever won the long-form content race used to get the first click; that click now goes to the AI answer

“actually it's not that it's harder to find your stuff becomes less relevant and that that's the part about SEO that I'm excited”
Eli Schwartz · 11:30

“up until AI overviews up until this entire concept of llm whoever won on that long form piece of content would get that first click…”
Eli Schwartz · 21:30

#ai-overviews#seo#funnel#discovery

Explainer1:10:00

Programmatic vs Editorial: Why TripAdvisor and Zillow Win

Eli explains programmatic SEO by contrast with editorial. TripAdvisor and Zillow merge many data sources (properties, cities, government records, UGC, photos) into one comprehensive page per entity, rather than hand-writing long-form content on each. For a home value or hotel rating, users don't want an essay, and editorial at that scale would be cost-prohibitive.

Programmatic = combining data sources into a page per entity; editorial = hand-written long-form per entity
TripAdvisor merged countries, properties, and UGC into pages that have dominated search for years
Zillow fused government data, comparison sales, photos, neighborhood and school data per property
Programmatic only works when there's both scale and a real user use case: scale for its own sake fails

“Trip Advisor did not write a piece of content on every single Hotel”
Eli Schwartz · 1:11:00

“SEO is the only channel to bring in traffic to each one of these Pages”
Eli Schwartz · 1:12:00

#programmatic-seo#editorial#tripadvisor#zillow

Story· 4

Story26:00

How Zapier's Legendary SEO Was Almost an Accident

Eli recounts advising Zapier ~10 years ago. People didn't search for Zapier because they didn't know it existed, but they did search for the tools they wanted to connect, like Gmail and Salesforce. The fix was building pages at scale showcasing every possible integration, creating a flywheel of discovery that became one of SEO's most-copied programmatic playbooks.

Nobody searched for Zapier, but everyone searched for the individual tools they wanted paired
The strategy was to build a marketing page for every possible integration (Gmail-to-Salesforce, etc.)
This created a flywheel: seeing one integration prompts users to look for the next
It became the canonical 'I want to imitate Zapier' example of programmatic SEO

“product which Zapped things together but no one needed because they didn't know it existed but people knew that they needed things to work together…”
Eli Schwartz · 26:30

#zapier#programmatic-seo#case-study#integrations

Story28:00

Tinder's SEO Win: Selling a Loneliness Solution, Not Dating

Working with Tinder, Eli found that all inbound came from the single word 'Tinder' and generic dating content wouldn't convert. User research reframed Tinder as a loneliness-solving product, which made it a local problem: someone new to a city is lonely. So they built programmatic 'online dating in [city]' pages, and this is why AI Overviews don't threaten it, the loneliness problem is mid/bottom-of-funnel.

Tinder had never done SEO; all inbound came from the branded term 'Tinder'
Reframing Tinder as a loneliness solution made it a local, city-by-city opportunity
They built programmatic 'online dating in [city]' pages worldwide (kept in beta)
AI Overviews can't solve loneliness, so the user still clicks Tinder's result: the value sits mid/bottom-of-funnel

“we discovered that Tinder is a loneliness solving problem loneliness solving solution so you're lonely you've gone to a new city you don't know anybody…”
Eli Schwartz · 29:30

“it's a loneliness solution problem that answers a Google search”
Eli Schwartz · 1:52:30

#tinder#user-research#programmatic-seo#case-study

Story42:30

The Gardening SaaS That Should Skip SEO for Trade Show Booths

A gardening SaaS was set on a $15,000/month, content-only SEO agency proposal. Eli asked how they actually got customers: booths at gardening shows costing $10,000 each. His advice was to spend the same budget on more booths, where visitors are in-market, try the tool live, and become real follow-up leads, rather than hoping content ranks for the wrong searches.

The agency proposal was $15,000/month, all content
Their existing customers came from in-person gardening show booths at ~$10,000 each
The same budget buys many booths with in-market users who try the tool and leave real leads
'There are searches for it but it's not the right searches' is the trap

“so I said instead of spending $115,000 on SEO you could get you can go to all these shows for the exact same budget”
Eli Schwartz · 43:00

#seo#channel-strategy#case-study#saas

Story53:30

How Quora Quadrupled Traffic by Un-Gating Its Answers

Eli advised Quora, which had never done SEO effectively. Two fixes drove a 4x traffic jump in three months: stop forcing users to log in before seeing answers (giving those answers to Google too), and build a categorized HTML sitemap of related questions so both bots and humans could navigate the whole site.

Quora gated answers behind login, hiding them from Google; showing them un-gated helped both users and Google
It had no internal navigation path, so Google couldn't crawl the site
A categorized HTML (not just XML) sitemap of related questions made all answers discoverable
Result: roughly 4x traffic within three months

“within 3 months they had quadruple traffic”
Eli Schwartz · 53:30

“yes they were going to be giving away answers for free but they were also going to be giving answers to Google”
Eli Schwartz · 54:00

#quora#sitemap#crawlability#case-study

Takeaway· 1

Takeaway59:30

Use AI for Product Descriptions, Not Blog Spam

Eli's rule for AI content: it's a tool, not a solution. Using AI to churn out worthless blog posts just produces the same worthless content for free instead of $50 on Upwork. But using AI to write product descriptions on an e-commerce category page is fine, because those pages rank on the products themselves, not the fluff content around them.

AI is a tool, not a solution: it only helps if the content is part of the user/product journey
Cheap AI blog spam is just the same worthless Fiverr/Upwork content at zero cost
E-commerce sites (JCPenney, Nordstrom, Macy's) rank on the products, not the category-page fluff
Google's own guidance: AI isn't the problem, unhelpful content is

“AI is a tool everyone says AI is a tool it's not a solution”
Eli Schwartz · 59:30

“using AI content to write more fluff content that's not necessary would just be a waste of time”
Eli Schwartz · 1:01:30

#ai-content#ecommerce#seo#content

Frameworks in this episode

Product-Led SEO

Programmatic vs Editorial Decision Rule

SEO Conversion-Metric Rigor

The Funnel-Position Defense Against AI Overviews

The SEO Go/No-Go Investment Test

Top-Down TAM SEO Forecast

Insights & moments

⚡Myth Buster· 3

Traffic Is a Vanity Metric If It Doesn't Convert