Citation-friendly pages
WEAKOnly one page (/top) matches the best-of, collection, or comparison shapes AI answer engines cite most. The single-meme explainers help, but roundup and 'best X memes' pages are almost entirely absent.
Your site barely shows up on Google for the searches your audience is running. Here's what's broken, and the 643k+ monthly searches you could rank for instead.
But they rank at positions 23 to 47, stuck on page 3 through 5, while a set of zero-difficulty terms in your exact wheelhouse (does he know meme, huzz meaning, locking in, the 41 meme, all at 49,500/mo) have no page targeting them at all.
MemePulse is a community meme platform: weekly meme battles readers vote in, a latest/top feed, and an archive of meme explainers ("meme stories") documenting formats like Geeked Patrick Star, Tung Tung Tung Sahur, and Aura Monster. The explainers are sorted into browsable categories (Reaction, Exploitable, Pop Culture, Sports, Slang) built to catch "what is X meme" and "X meme meaning" searches from a Gen Z, internet-culture audience.
Registered members post memes, upload media, and comment. The technical base is clean: schema on all 15 pages we fetched, an H1 on every one, meta descriptions present, and an llms.txt manifest already live.
The problem isn't the plumbing, it's what you're pointing it at. You've already proven the "meme meaning" thesis works: jestermaxxing, happy patrick, thank you chainsaw man, and geeked patrick all rank, some for keywords pulling 8,000 to 12,000 searches a month.
Two moves matter: rebuild the handful of articles already ranking so they climb off page 3, and publish against the KD-0 gaps before knowyourmeme locks them up.
5 structural issues. 4 sitemap bloat flags.
Technically the site is tidy. All 15 fetched pages return 200 with schema present, there are 0 noindex pages, 0 canonical issues, 0 duplicate titles or metas, image alt coverage is 75%, and every URL carries a fresh lastmod. The /meme/ explainer archive is a genuine asset: 170 pages targeting real high-volume format queries like jestermaxxing (12,100/mo) and happy patrick meme (9,900/mo).
Structurally, though, the site is hollow. Of 4,444 sitemap URLs, 3,305 (74%) are machine-translated /es/, /fr/, and /pt-br/ mirrors that hold 0 ranked keywords, and another 847 are numbered /media/ viewer pages that also rank for nothing. The entire search footprint sits in one bucket: /meme/ carries 52 of the site's 54 ranked keywords, and the whole ranked set is worth an estimated 186.25 in traffic. Everything the brief sells (voting, battles, timelines, media uploads) is invisible in search, and 50 of the top-54 head keywords sit at position 21 or worse. The bones are clean; the body is mostly programmatic filler wrapped around a small archive that ranks too deep to earn clicks.
3,305 URLs live under /es/, /fr/, and /pt-br/, which is 74% of the 4,444-URL sitemap. All three are machine-translated clones of the same English browse shells: /es, /es/latest, and /es/media mirror the English tree one-for-one, and together they hold 0 ranked keywords. Google's 2024 Helpful Content and spam updates devalue mass low-quality programmatic pages, and at this volume the translations, not the 170 explainer pages that actually rank, define the domain's quality profile. Because each language duplicates thin browse pages rather than the explainer text, the site multiplies near-duplicate low-value URLs by three.
1. Pull the non-indexed URL list for each language prefix from Search Console.
2. If you are not actively serving Spanish, French, and Portuguese audiences, noindex all three trees, remove them from the sitemap, and 410 the thin browse shells.
3. If international traffic is a real goal, keep only human-reviewed translations of the /meme/ explainers, add reciprocal hreflang tags, and drop the translated browse, media, and vote shells.
4. Re-scan the sitemap and confirm it falls from 4,444 toward the ~1,100 URLs that carry unique value.
847 numbered pages sit under /media/ (/media/1, /media/11, and so on), each a single-image viewer with roughly 220 words of templated text and 0 ranked keywords. The same media tree is cloned inside every language mirror, so the true count of thin image pages is far higher. These pages carry no standalone search intent: nobody queries an image ID, and the body is boilerplate navigation rather than unique description. They inflate the crawlable surface, spend crawl budget that should reach the explainer archive, and add to the low-value page ratio that quality classifiers measure. A meme image belongs to its parent explainer, not a separate indexable URL.
1. Add noindex to all /media/<id> viewer pages across every language.
2. Set a canonical from each media page to its parent /meme/ explainer.
3. Remove the numbered media URLs from the sitemap so only explainer and category pages remain.
4. Keep the viewer live for on-site UX, but let the parent explainer be the one indexable home for the image.
The search footprint lives in one bucket. /meme/ holds 170 pages and 52 of the site's 54 ranked keywords; /categories/ adds 1 and /latest/ adds 1. Every other prefix earns nothing: /vote/ (20 pages, 0 keywords), /timeline/ (80 pages, 0), /media/ (847 pages, 0), plus /winners/, /top/, and the three language trees (3,305 pages, 0). The community product the brief is built around, weekly battles, voting, media uploads, and timelines, is completely invisible in search. Total estimated traffic value across the ranked sample is 186.25. Growth cannot come from the community shell, because Google does not rank interactive utility pages for informational meme queries.
1. Treat the /meme/ explainers as the only SEO surface and set the content roadmap there.
2. Keep vote, timeline, and media pages for logged-in engagement, but noindex them so they stop diluting the quality signal.
3. Expand the explainer archive against the pillar themes already surfaced (reaction memes, meme meaning, brainrot memes).
4. Internally link every new explainer from its category page to concentrate authority on the pages that can rank.
The indexable browse pages are thin and near-identical. The homepage runs 540 words, /latest 309, /media 220, and /top just 200, and every fetched page repeats the same H2 navigation block: Explore MemePulse, Meme media, Meme categories, Account and community, MemePulse. That boilerplate consumes most of the unique-text budget on each shell, leaving little for Google to distinguish one page from the next. /top is worse: it indexes the empty state 'No featured memes yet for this range,' publishing a blank page.
1. Add a unique 100-plus-word intro to each indexable browse and category page describing what it collects and why.
2. Move the repeated Explore MemePulse navigation block into a footer component so it is not the page's primary heading structure.
3. Noindex empty-state pages like /top until they hold real entries.
4. Give each category page distinct supporting copy so it targets its own 'X memes' query instead of echoing the others.
The explainers target the right terms but rank too deep to earn clicks. jestermaxxing (12,100/mo) sits at position 47, happy patrick meme (9,900) at 34, thank you chainsaw man (8,100) at 28, and always has been (6,600) at 46. Of the top-54 head sample, 50 keywords sit at position 21 or worse and only 1 reaches page 1, which is why the whole ranked set is worth an estimated 186.25. The cause is depth and authority. knowyourmeme.com ranks for 465,282 keywords at an average position of 8.8 with 91,008 referring domains; the explainers here are shorter and the domain carries almost no link equity, so it loses format-explainer SERPs it already appears in.
1. Rank the /meme/ pages by search volume and rewrite the top 20 with deeper origin, spread, usage, and variation sections that beat the incumbent entry.
2. Capture the two striking-distance terms first: 'why patrick so geeked up' (position 10) and 'thank you chainsaw man panel' (position 14) need only small on-page gains.
3. Add example images and dated first-sighting citations, the elements format-explainer queries reward.
4. Earn a handful of referring domains by being the first cited source on brand-new formats.
3 pages analysed for AI-writing patterns.
Scored 52/100 across 3 pages. The writing has some shape but several tells per page (14 flagged) suggest AI assistance without an editor's pass. Cleanup is straightforward, not a rewrite.
“un anuncio sencillo se vuelve señal para imitaciones, edición surrealista y cruces brainrot”
Rule-of-three list ending a sentence is classic LLM cadence for padding a definition with three parallel nouns.
“explicaciones circulares u obviedades vendidas como oportunidades”
Second rule-of-three enumeration in a short paragraph, the LLM habit of listing three near-synonymous items instead of one.
“El discurso recortado se vuelve una tautología accidental”
Over-tidy explainer diction that abstracts a joke into a definitional claim, a common LLM register in short reference blurbs.
“YouTube — GLITCH, final Remember de The Amazing Digital Circus”
Em dash used as a separator in the source list; a human editor would use a colon or period.
“X — @forever_gf618, primera cuenta con el meme hallada en estas fuentes”
Another unedited em dash separator; the sources block stacks five of these in a row.
“Bluesky — @avidgladiator, variante temprana con personaje redibujado”
Repeats the em-dash-as-label pattern; five em dashes total across the citations trips the hard floor.
“Además, es una fórmula modular”
Paragraph-internal 'Además' (Moreover/Additionally) is filler an editor would cut.
“presentar una respuesta absurda como si fuera un gran avance, unir fandoms sin relación o revelar un objeto entre las manos de Kaufmo”
Rule-of-three parallel list, a common LLM scaffolding shape.
“la ropa, la cara y hasta la identidad de Kaufmo pasaron a formar parte del remate”
Second rule-of-three enumeration in a short span reinforces the formulaic rhythm.
“Puede ser un chiste de coqueteo, de oversharing o un redraw sin perder la idea base.”
Rule-of-three list stacked to manufacture depth. Classic LLM cadence when explaining flexibility of a format.
“química incómoda en la oficina, enseñar algo que nadie pidió ver o actuar con demasiada confianza”
Second three-item parallel construction in adjacent paragraph. The doubled rule-of-three is a strong AI rhythm tell.
“La foto funciona porque ya parece una escena”
'funciona' recurs as section header, topic sentence, and later 'funciona muy bien', giving the formulaic works-because explainer voice.
“Esa tensión incorporada vuelve al formato muy fácil de reescribir.”
Abstract cause-effect filler ('built-in tension makes it easy to rewrite') with no concrete detail. Generic LLM analysis padding.
“Among Us — Sus”
Unedited em-dash in the related-meme heading; only one on the page, but a machine-set separator a human would style differently.
100% of sampled rankings come from searches that don't contain the brand name. That's traffic SEO earned, not reputation.
Already ranking close to page 1. These move first: a sharper title, a content refresh, a few internal links.
SERP FEATURES · 39 keywords show People Also Ask · 16 show AI Overviews
Rough equivalent ad spend based on current click prices. Directional, not an invoice.
Keywords where a competitor's specific page outranks yours, sorted by what the gap is worth. Each row is one page to build or upgrade.
NO PAGE = you have nothing ranking for that keyword at all
Under 30 characters. Likely missing keyword context.
Under 120 characters. Wasting SERP real estate.
Pages under 300 words. May lack enough depth for competitive terms.
Based on the 15 pages that returned HTTP 200 out of 15 fetched. Not a full-site crawl.
Click depth and inlink counts are computed over the 15 pages we fetched, not a full crawl. Pages outside the sample are not represented in the graph.
Compared the first 15 sitemap URLs against the top-500 ranked-keyword URL set.
Classic rankings and AI citations are now separate games. This checks whether the site has pages and entity signals worth retrieving.
MemePulse appears in 16 AI Overviews while knowyourmeme, imgflip, and x.com each hit 100, so answer engines already know the topic and mostly cite competitors. The base signals are there (schema on every page, an llms.txt manifest live), but the two levers that get a site cited are weak: the homepage doesn't clearly state what MemePulse is, and only one page is the best-of/collection format AI answers pull from. Fix those and the AIO share should climb.
Only one page (/top) matches the best-of, collection, or comparison shapes AI answer engines cite most. The single-meme explainers help, but roundup and 'best X memes' pages are almost entirely absent.
The homepage title is 'Weekly Meme Voting and Latest Memes' and the H1 is 'New memes'. Neither states the category (meme format archive), the audience, or the proof, so AI systems can't cleanly identify the entity.
Schema is present on all 15 fetched pages and the meme explainers already carry origin/meaning structure, which is the format AI Overviews quote. This is the strongest input you have.
No Wikipedia entity found and no video or social proof in the data. For a meme brand, short YouTube format-explainer clips are a natural, low-cost entity signal and a medium-impact move, not a first bottleneck.
memepulse.net shows in 16 AI Overviews versus 100 each for the three main competitors, and is absent on 10 terms where a competitor is cited, including '6 7' and '6'7' via knowyourmeme.
No evidence of any AI-answer visibility tracking. Sources in AI answers reshuffle often, so check your citation share on the top meme-meaning terms weekly or monthly.
AI answers and browse searchers both favor roundup pages, and you have only one (/top). Build 'best TikTok reaction memes', 'Italian brainrot memes explained', and a Gen Z slang dictionary from your existing keyword list. These are the page types that get cited.
Change the title and H1 so they name what MemePulse is (a meme format archive and community), who it serves (meme creators and internet-culture followers), and the proof (weekly battles, growing explainer archive). AI systems cite homepages and clear brand profiles; 'New memes' gives them nothing to anchor to.
On terms like the 6-7 meme where knowyourmeme is cited and you are not, your explainer format directly competes. Expand those articles with a tight definition block up top (the format, the origin, the meaning in one or two sentences) so an answer engine can lift it cleanly.
Highest-leverage first. Keywords with a ready article on top.
Own two pillars first: "meme meaning / meme explained" and "reaction memes." You already rank for the meaning cluster (jestermaxxing, happy patrick, chainsaw man, geeked patrick), so the fastest path is depth, not breadth: turn each explainer into the definitive page for that format, then link them under hub pages for brainrot, reaction, and slang. The keyword list hands you the hubs directly (brainrot memes at 18,100/mo, gen-z slang at 49,500/mo), and dozens of low-difficulty leaf terms (huzz meaning, does he know meme, 67 meme meaning, locking in) that map one-to-one onto single articles. Publish the KD-0 terms first while they're uncontested.
One keyword in, one published article out, every day. We'll ship against the 30-keyword list above.
REAL DATA ABOVE · 5 QUICK WINS, 8 ARTICLE IDEAS, 30 KEYWORDS FOR MEMEPULSE