Lenny's PodcastThe ultimate guide to SEO | Ethan Smith (Graphite)
CHAPTERS
- 0:00 – 7:53
Ethan Smith’s path from UX research to SEO, and what Graphite does
Ethan explains how he moved from UX/user research into SEO out of necessity, then grew into product/growth roles where SEO required core product changes. He shares how Graphite formed from consulting demand and why their differentiator is execution—not just strategy.
- •Started in UX research/design; learned SEO to solve expensive acquisition
- •Early SEO learning came from a strong consultant plus years of testing
- •Belief: great SEO often requires product changes, not just marketing tweaks
- •Graphite grew from inbound projects (MasterClass, Honey, Ticketmaster)
- •Agency focus: SEO + content strategy with an emphasis on building/shipping
- 7:53 – 10:02
The biggest SEO myths: “technical audits” and over-fixating on technical SEO
Ethan argues that traditional technical audits often produce bug lists that don’t drive growth. He reframes technical SEO as supportive infrastructure and claims most growth comes from programmatic and especially editorial SEO.
- •Technical audits often become unprioritized bug reports, not growth plans
- •Technical SEO is commonly misunderstood and over-emphasized
- •Growth tends to come from programmatic and editorial SEO
- •Editorial SEO (human-written content) is underappreciated
- •Technical work should be driven by strategy, not generic checklists
- 10:02 – 17:22
When SEO is worth investing in: authority signals and market size
Ethan shares two core gates for SEO: a sufficiently large addressable market and enough authority/traction to be credible in Google’s eyes. He offers practical heuristics (non-SEO traffic and referring domains) and explains product vs. audience competitors.
- •Two decision factors: addressable market size + existing authority/traction
- •Heuristic: ~1,000 non-search visits/day suggests enough brand credibility
- •Heuristic: ~1,000 referring domains makes ranking meaningfully easier
- •Use Similarweb for traffic (free) and Ahrefs/Semrush for referring domains
- •Differentiate product competitors vs. audience competitors (e.g., Robinhood vs. Investopedia)
- 17:22 – 18:36
Impact vs. scope: why UGC/SKUs make SEO easier, and how to choose a focus
The discussion expands from “should we do SEO” to “is it efficient for us right now.” Ethan explains how UGC, inventory/SKUs, and database-driven content reduce the scope of competing, and how to compare SEO against paid, social, and other channels.
- •Use an impact vs. scope lens across growth channels
- •UGC-heavy sites (e.g., Zillow/Glassdoor) reduce SEO scope via scalable pages
- •No UGC doesn’t disqualify you—editorial SEO can still win
- •Estimate value by competitor traffic × your conversion rate (vs. other channels)
- •Timing depends on cost/opportunity tradeoffs, not just SEO best practices
- 18:36 – 23:29
The three buckets of SEO: programmatic, editorial, and technical
Ethan defines the three major SEO types and what they look like in practice. He also notes how editorial SEO has grown to drive more traffic than programmatic for many companies, while technical SEO underpins both.
- •Programmatic SEO: auto-generated pages from databases (items + categories)
- •Editorial SEO: human-written articles, guides, listicles
- •Technical SEO: infrastructure like internal links, redirects, tags, speed
- •Programmatic is cheaper per page at scale; editorial has marginal cost per page
- •For most non-commerce companies, editorial is the primary opportunity
- 23:29 – 28:14
Reverse-engineering Google: choosing the right page type and page requirements
Ethan outlines a practical method: start from target queries, inspect what Google ranks, and infer page types and content components that match intent. He stresses that the goal isn’t copying, but using patterns to build a better page that wins on engagement.
- •Determine page type by searching target terms and observing SERP patterns
- •Example: MasterClass “chapter transcript” pages won’t rank if SERP wants recipes/articles
- •SERPs can show fractured intent (e.g., listicle vs. category page)
- •Use competitors’ page components (maps, phone numbers, etc.) as intent signals
- •Optimize for user utility to improve engagement and resilience to updates
- 28:14 – 29:48
Site engagement: the ranking signals many teams ignore
Ethan breaks down engagement into SERP behavior (CTR, pogo-sticking) and on-page behavior (time/continuation). He argues that better utility drives better engagement, which can improve rankings quickly—not just long-term robustness.
- •Engagement includes CTR at your rank position on the SERP
- •Bad signal: clicking your result then returning to Google for another result
- •Google can infer post-click behavior via Chrome/Android signals
- •Goal: users find the answer and end the session or continue deeply on your site
- •Better engagement can lift rankings immediately, not only over months
- 29:48 – 31:41
Indexing strategy and the shift from keyword pages to topic pages
Ethan explains why “one keyword = one page” is outdated and can create low-quality, duplicative content. Modern SEO clusters many keyword variants into topic pages and selectively indexes pages based on quality thresholds.
- •Index high-quality item/product pages; avoid thin/out-of-stock/empty pages
- •Old approach: millions of near-duplicate category pages for keyword variants
- •Modern approach: cluster keywords into topic pages Google understands semantically
- •A single page may rank for ~200–2,000 keywords within a topic
- •Topic thinking prevents waste and aligns content with broader intent coverage
- 31:41 – 36:42
Topics vs. keywords in practice: clustering, subtopics, and prioritization
Using examples like “butter lettuce” and “five-year plan,” Ethan shows how topics contain many keyword themes (templates, examples, benefits, comparisons). He shares how to decide whether to keep content unified or split into separate pages based on SERP overlap.
- •Topic pages must cover key subthemes to be considered comprehensive
- •Gaps reduce content quality signals and cause users to bounce
- •MasterClass example: outranking bigger sites by covering missing subtopics (e.g., health benefits)
- •BetterUp example: one page ranking for ~1,000 keywords by including templates/examples/variants
- •Decide split vs. single page by comparing SERP result overlap across queries
- 36:42 – 40:14
Competitor mining + essential tools for building an SEO strategy
Ethan describes how to bootstrap strategy by extracting competitors’ top pages/keywords and building a prioritized topic list. He recommends a small core toolkit (Search Console, Clearscope, Ahrefs/Semrush) and explains where Screaming Frog fits.
- •Mine competitor sites via Ahrefs/Semrush: top pages, keywords, patterns, traffic proxies
- •Build a large topic list and prioritize by volume, competitiveness, and conversion intent
- •Google Search Console: free, real query data, SERP tracking, performance monitoring
- •Clearscope: best-in-class content scoring and gap discovery for outlines/subtopics
- •Screaming Frog: crawling, error discovery, and internal link coverage audits
- 40:14 – 45:03
How long SEO takes—and the leading indicators to watch
Ethan sets expectations: most companies see meaningful results in ~6–9 months, but high-authority domains can see changes in days or weeks. He suggests watching early ranking movement (e.g., reaching position ~20 then climbing) as proof of traction before scaling.
- •Typical timeline: ~3 months to build/launch + ~3–6 months to see ranking growth
- •High authority brands can see impact extremely fast (days to weeks)
- •Google tests newer/low-authority sites on less competitive terms first
- •Leading indicator: pages start ranking (even with low clicks) and move upward steadily
- •Avoid false negatives: lack of rank often means poor implementation or insufficient authority
- 45:03 – 57:15
Hiring for SEO: what to look for in programmatic vs. editorial roles
Ethan differentiates hiring needs: programmatic SEO is specialized and best staffed by technical growth operators with prior experience, while editorial SEO can be learned with the right workflows. He outlines what each role does day-to-day and how to run SEO experiments.
- •Programmatic SEO hires: technical growth profile; ideally has shipped it before
- •Recruiting hack: find lesser-known sites with unusually high % of traffic from SEO and hire their operators
- •Programmatic day-to-day: competitor analysis, requirements, keyword clustering, partnering with engineers, measurement
- •Testing approaches: A/B or sequential; some domain-wide effects can’t be cleanly tested
- •Editorial workflow: SEO manager selects topics + outlines; domain-expert writers draft; then publish and iterate
- 57:15 – 1:06:09
Scaling editorial SEO: team structure, content velocity, and page types
Ethan shares how many articles to produce early vs. after proof of traction, and what the minimum viable team looks like. He also summarizes the dominant page types across programmatic, editorial, and special cases like calculators and integrations.
- •Typical starting cadence: ~10–25 articles/month (often ~25) to validate traction
- •Scale after leading indicators; some companies ramp to 100+ articles/month
- •MVP team: one SEO strategy person + one writer (SEO managers often don’t write)
- •Core page types: category pages, item pages, articles, listicles, guides, PM/landing pages, support pages
- •Special page types: calculators (e.g., mortgage/BMI), Zapier-style integration pages
- 1:06:09 – 1:08:38
Getting SEO shipped: overcoming roadblocks, org design, and executive buy-in
Ethan explains the “dirty secret” that many SEO recommendations never launch due to resourcing constraints and misalignment with product teams. He advocates for product-led SEO, executive sponsorship, and prioritization systems that treat SEO work as real product work.
- •Most SEO work fails because it never ships—not because the strategy is wrong
- •SEO is often outside product org despite requiring product/engineering changes
- •Product-led SEO framing helps integrate work into roadmaps
- •Executive buy-in can unblock resources for high-leverage initiatives
- •SEO’s perceived “mystery” makes it harder to prioritize vs. clearer channels
- 1:08:38 – 1:16:36
Topical & domain authority: how Google decides what you’re “known for”
Ethan reframes authority as multi-input (links, branded search, non-SEO traffic, shares) and importantly topical, not just domain-wide. He gives examples of how a viral post can lift adjacent topics across a domain, and how to expand outward from existing topical strength.
- •Authority evolved beyond PageRank: includes traffic signals, branded search, and social/shares
- •Authority is topical: Google evaluates what your site is disproportionately strong in
- •Viral-content example: “Jell-O shots” boosted rankings for adjacent alcohol topics site-wide
- •MasterClass strategy: start with Gordon Ramsay-adjacent topics before broader food terms
- •Topics relate via adjacency graphs that Google likely builds algorithmically from co-occurrence data
- 1:16:36 – 1:24:28
The power of internal links: the biggest ‘quick win’ lever
Ethan argues internal linking is one of the most under-optimized, high-impact SEO levers, often producing large lifts. He explains why many sites create “power curves” where only a few pages get most internal links, and why body links matter more than nav/footer links.
- •Internal links help Google crawl, refresh, and assign importance to pages
- •Ethan found link-count correlated with traffic; adding links produced ~300% lift in one case
- •Most sites have a skewed distribution: few pages get most links; many get near-zero
- •Common link systems (popular/recent/related/faceted) fail to distribute links broadly
- •Body links > nav/footer; place link modules above the footer for stronger impact
- 1:24:28 – 1:28:29
AI and SEO content: why generation is risky, and where AI actually helps
Ethan cautions against AI-generated articles because they can be fluent but factually wrong—dangerous in domains like health and finance. He argues AI is more valuable upstream (clustering, outlining, subtopic extraction) and can be safe for writing only when grounded in structured data.
- •GPT-style generation can hallucinate facts (e.g., incorrect product claims like “oxide-free”)
- •Fluent writing isn’t the same as useful, truthful content
- •Best use: automate research tasks—topic clustering, subtopics, outlines, authority estimation
- •Safe generation can work when constrained by structured data (e.g., sports play-by-play, structured clinical trial info)
- •Recommended workflow: AI assists structure; humans with domain expertise write final content
- 1:28:29 – 1:30:59
Final ‘getting started’ checklist: validate market, validate authority, compare alternatives
Ethan closes with the first steps a listener should take: quantify the SEO market via competitor traffic, assess whether you can compete via authority signals, then compare SEO ROI against other growth options. Lenny wraps with where to find Ethan and Graphite.
- •Step 1: size the addressable market via product + audience competitors and their traffic
- •Step 2: evaluate your authority (non-SEO traffic, referring domains, traction)
- •Step 3: prioritize SEO vs. other channels based on scope, cost, and expected return
- •Graphite contact: graphitehq.com and ethan@graphitehq.com
- •Theme recap: execution and shipping matter as much as strategy
