Before we run the audit, we need to make sure we're asking the right questions about the right competitors to the right buyers. This document presents what we've learned about Paramount+'s market — your job is to tell us what we got right, what we got wrong, and what we missed.
Before we measure how often AI assistants cite Paramount+ in streaming recommendations, these three signals tell us whether AI crawlers can reach, read, and trust the site's own pages — the precondition for being cited at all.
AI assistants are quietly becoming the front door to streaming decisions. When a household asks "what's the best streaming service for live NFL without cable" or "is Paramount+ worth it now," the answer increasingly comes from an AI engine synthesizing sources — not from a Google results page the household scrolls itself. In a subscription video streaming category this crowded and this churn-sensitive, the services that AI platforms learn to cite first build a compounding advantage: early citations become self-reinforcing as engines learn which domains to trust. Paramount+ enters this shift as an established, enterprise-scale service — the question is whether its own pages are shaping those AI answers, or whether third parties are.
This Foundation Review is what we validate together before the audit runs. It lays out three things: the competitive set that shapes which head-to-head streaming queries we test, the subscriber archetypes that determine how buyer intent is phrased, and the Layer 1 technical baseline that determines whether AI crawlers can even access Paramount+'s content. None of this is the audit itself — it's the shared map we confirm so the audit asks the right questions of the right buyers against the right rivals.
The validation call is a working session with real stakes. Two kinds of decisions get made: input validation — are the right competitors in the right tiers, and are these the right subscriber archetypes? — and engineering triage — which technical fixes can start before results come back? Your answers directly set how the buyer query set is partitioned and weighted across the AI platforms we test. The specific items to decide are in the Pre-Call Checklist near the end of this document.
Three quick notes on what this is, what we need from you, and how to read the confidence badges throughout.
What this is This is the foundation for a Generative Engine Optimization (GEO) audit — a measurement of how AI assistants represent Paramount+ when subscription video streaming buyers ask them for recommendations. Everything here is an input we're validating, not a result. The audit itself comes after this document is confirmed.
What we need from you Read each section as a claim we're making about your market. Where we're right, confirm it. Where we're wrong, correct it. Where we've missed something — a competitor, a subscriber type, a pain point — tell us. The purple boxes throughout flag the specific questions where your answer changes how the audit is built.
Confidence badges Every entity carries a confidence badge. High means it's sourced from scraped site data or corroborated across multiple review sources. Medium means it's inferred or single-sourced and warrants a closer look at the call. Focus your attention on the medium-confidence items first.
The baseline profile we built from the Paramount+ site and public sources. Confirm the essentials before we build queries on top of them.
→ Paramount+ spans on-demand originals, live TV and news, live sports, and kids content under a single subscription — do buyers actually evaluate it as one service, or do the live-sports, prestige-drama, and kids audiences run genuinely separate consideration sets? If they're separate, we partition the audit into distinct query clusters per vertical instead of one blended Paramount+ query pool — the single biggest structural decision for the query set.
5 subscriber archetypes: 3 decision-makers, 1 evaluator, 1 influencer. These archetypes drive how the audit phrases buyer intent — each one searches differently, so each gets its own query cluster.
Critical review area Personas are the highest-leverage input in this document — they determine the intent behind every query we run. This is a consumer (B2C) adaptation: "personas" are subscriber archetypes and household decision-makers, not a B2B buying committee, and "seniority" encodes household subscription authority (C-Suite = the primary budget-holder). If an archetype is wrong or redundant, the audit spends its query budget simulating a buyer who doesn't exist.
Data sourcing note Names, roles, seniority, veto power, and technical level are drawn from the knowledge graph (mostly from review mining). The role descriptions, buying jobs, and query focus areas below are synthesized from those fields to show how we'd model each buyer — treat the synthesized lines as our interpretation to confirm or correct.
→ Does Marcus keep Paramount+ year-round, or churn seasonally around the sports calendar the way Tyler Brooks does? If he's a seasonal subscriber, his acquisition-stage sports queries overlap the budget-churner's set and we consolidate them rather than double-counting.
→ Does Diane evaluate the Showtime/Premium tier specifically, or Paramount+ broadly? If the Showtime paywall is her real decision point, we route her queries to Premium-tier and Showtime-content comparisons rather than base-plan queries — which changes which tier the audit tests her intent against.
→ Is Priya's subscribe trigger Nickelodeon specifically — a direct kids head-to-head with Disney+ — or general family viewing? If it's Nickelodeon, we add kids-safety and Disney+ comparison queries; if it's general family, kids becomes one facet of a broader household bundle decision rather than its own cluster.
→ Building on the company-profile question: does Tyler's churn track pricing and free-trial changes, or the sports/premiere calendar the way Marcus's does? The distinction determines whether cancellation queries cluster with price-and-value intent or with content-availability intent.
→ This persona is llm-inferred, not review-sourced. Does Jordan search distinctly from Diane Whitfield, or do both simply type franchise/title-availability queries? If their queries are identical, we merge them and reallocate the budget — keeping a redundant persona dilutes visibility-by-buyer reporting.
Missing subscriber types? These archetypes sometimes show up in streaming decisions — do they show up in yours? (1) The bundle-acquired subscriber who got Paramount+ through Walmart+, an Amazon Prime channel, or a mobile carrier, and searches "is Paramount+ included with…" rather than evaluating it directly. (2) The international / expat sports viewer chasing specific soccer or cricket rights. (3) The older CBS-loyal broadcast viewer who wants live local CBS, news, and daytime shows more than originals. Each would warrant its own query cluster. Who else shows up in your subscriber base?
5 primary + 4 secondary competitors identified. Tier assignments determine which rivals get head-to-head query coverage in the audit.
Why tiers matter Each primary competitor gets roughly 6–8 head-to-head queries — the "Paramount+ vs Peacock for sports," "best streaming for prestige drama," and "Paramount+ vs Disney+ for kids" style prompts that test direct differentiation. Getting these tiers right determines which ~30–40 queries test direct competition versus category awareness. We're least certain about Apple TV+, placed as primary at medium confidence — it overlaps with Paramount+ only on the "subscribe for one show" churn pattern, so if it rarely appears in real deals, moving it to secondary reallocates that head-to-head budget.
→ Three things to confirm: (1) Is Apple TV+ a real primary, or should it drop to secondary — is your true head-to-head just Peacock, Netflix, Max, and Disney+? (2) Starz and ESPN are llm-inferred secondary picks — do they actually surface in real Paramount+ consideration? (3) What's missing — free ad-supported rivals (Tubi, or your own Pluto TV), live-TV bundlers (YouTube TV, Fubo, Sling), and how the pending Max merger reshapes a rival that's about to become a sibling? Any competitor you'd cut or add here changes the head-to-head query set directly.
11 buyer-level capabilities mapped — 4 strong, 6 moderate, 1 weak. These determine which capability queries the audit tests and where it presses hardest.
Watch live NFL on CBS, UEFA Champions League soccer, and UFC without a cable package.
Shows I can only get here — Star Trek, the Yellowstone/Taylor Sheridan universe, Showtime originals.
Stream my local CBS station, live news, and current-season network shows the day after they air.
Nickelodeon and safe kids shows with a separate profile I can trust my child to use.
Is there actually enough to watch here, or do I run out after the one show I came for?
Does the app actually work — or does it buffer, freeze, and kick me back to the home screen mid-episode?
Is it worth the monthly price, especially now that they raised it and killed the free trial?
Will it run well on my smart TV, Roku, phone, and game console — not just technically "supported"?
How many ads am I going to sit through on the cheaper plan before I give up and pay more?
Can I download shows for a flight and stream on three screens at once for my family?
Can I get it cheaper through a bundle, a student discount, or an annual plan instead of paying full price?
Which strengths do you lead with? The audit tests all 11 capabilities, but competitive-differentiation queries will emphasize 3. Four are rated Strong:
• Live Sports Rights & Streaming — live NFL, Champions League, UFC without cable
• Original & Exclusive Series and Films — Star Trek, the Sheridan universe, Showtime
• Live TV, Local Stations & News — local CBS, live news, next-day network
• Kids & Family Content with Kid Profiles — Nickelodeon with trusted kid profiles
Which of these best represents where Paramount+ actually wins subscribers? Your top 3 become the differentiation emphasis in the query set.
→ Two calibration questions: (1) Are the ratings honest relative to named rivals — is Streaming Reliability genuinely your weakest capability versus Netflix and Disney+ on the same smart TV, and is On-Demand Library Breadth only "moderate" against Netflix and Max? (2) Are any of these merge candidates — Device Compatibility and App Reliability overlap heavily in the reviews — and is anything missing (4K/HDR quality, content discovery/recommendations, simultaneous-stream limits)? Mis-rating the weakest feature points the vulnerability analysis at the wrong target.
10 pain points: 4 high, 6 medium severity. Buyer language here is how the audit phrases the frustration-driven queries subscribers actually type.
→ Two of the high/medium pains lean on inference: local sports blackouts is rated High but is llm-inferred, and content removed mid-rewatch is llm-inferred Medium — do both surface strongly enough in real subscriber complaints to anchor vulnerability queries? And what's missing: the password-sharing / extra-member crackdown and fees, login/authentication friction across TV providers and devices, and poor content discovery ("I can't find anything to watch") are common streaming pains — do they show up for Paramount+? Getting severity wrong changes which frustrations the audit tests first.
The technical state of the site as an AI crawler sees it. These are handoffs for your engineering team — fixes that improve the baseline before the audit even measures citation visibility.
Engineering — start now The good news first: robots.txt is confirmed open — GPTBot, ClaudeBot, PerplexityBot, Google-Extended and Googlebot are all allowed, so this is a rendering problem, not an access block. But the rendering problem is severe. The Help Center serves a JavaScript loading shell with zero crawlable support content (critical), core catalog, browse, and signup pages render as near-empty shells, and three marquee show-hub routes return HTTP 500 to non-browser crawlers. Engineering should start on the show-hub 500 (a 1–3 day fix) and the Help Center / catalog server-side rendering immediately — these don't require the validation call.
What we found: Every page fetched on the help.paramountplus.com subdomain (the Help Center home plus 6 individual articles: Supported Devices, live-TV/local-CBS availability, profiles & kid profiles, "Have our Plans Changed?", and Xfinity/Xumo streaming) returned only a Salesforce Experience Cloud loading shell — visible text was limited to "Paramount+ Help Center," "Sorry to interrupt," a "CSS Error" notice, and a "Refresh" link. No article title, headings, or body text rendered. The article body is injected client-side and is absent from the server-rendered output a crawler receives.
Why it matters: Help/FAQ content ("how do I cancel," "what devices are supported," "is my local CBS included," "did the price change") is among the most-cited material in AI answer engines for subscription-service questions, because it's authoritative and answers buyer intent directly. If the entire Help Center is invisible to non-JavaScript crawlers, AI assistants answer these high-intent questions from third-party sites (Reddit, cabletv.com, news articles) instead of Paramount+'s own words.
Recommended fix: Server-side render or pre-render the Help Center article content (or expose it via a crawlable static/SSR mirror) so the full question-and-answer body is present in the initial HTML. At minimum, ensure each article returns its title and answer text without requiring JavaScript execution. Validate by fetching several article URLs with JavaScript disabled and confirming the answer text is present.
What we found: The primary catalog pages — /shows/, /movies/, /collections/, /browse/originals/ and a UFC collection page — returned output that was ~95% navigation and footer boilerplate with essentially no body: no show/movie grid, titles, descriptions, or collection copy. The signup flow page /account/signup/createaccount/ returned only footer/legal nav with no plan selector or pricing table. Separately, several content routes returned a generic marketing template instead of their own content (e.g., /news/ and /live-tv/stream/cbssportshq/ served the homepage "A Mountain of Entertainment" shell).
Why it matters: These pages are the crawlable representation of the library's breadth and the plan/pricing offer — the exact things buyers ask AI assistants about ("what's on Paramount+," "is it worth it," "what does it cost"). When the catalog grid and plan table are injected client-side, a non-JS crawler sees an empty page, so the depth of the library and the value proposition are invisible at the source.
Recommended fix: Server-side render or pre-render the catalog/browse grids and the plan-selection/pricing content so titles, descriptions, and plan details appear in the initial HTML. Ensure content routes (/news/, live-channel pages) return their own server-rendered content rather than falling back to the generic marketing template. Verify with JavaScript disabled.
What we found: Three individual show/franchise hub URLs returned HTTP 500 Internal Server Error to our automated fetch: /shows/tulsa-king/, /shows/star_trek_animated/, and /shows/uefa-champions-league/. These are marquee-title landing pages (Tulsa King, Star Trek, UEFA Champions League) that a crawler or AI engine would reach directly from search. A sibling episode-listing route (/shows/the_price_is_right/episodes/) did render, so the failure is specific to the top-level show-hub route, not the whole /shows/ tree.
Why it matters: A 500 response means AI crawlers (GPTBot, ClaudeBot, PerplexityBot, Googlebot) get nothing at all for the client's most valuable, search-visible originals and sports franchises — the pages most likely to surface for "where to watch Tulsa King" or "stream Champions League." Persistent 5xx errors also erode crawl trust and can cause engines to deprioritize the URL entirely.
Recommended fix: Reproduce the 500 on these show-hub routes with a plain (non-browser) user-agent and from a datacenter IP, fix the server error, and confirm a 200 with rendered content. Check server logs for the error class and whether it correlates with bot user-agents or missing JS-only parameters. Add these URLs to monitoring so future 5xx regressions on show hubs are caught.
What we found: https://www.paramountplus.com/sitemap.xml returned HTTP 404, as did /sitemap_index.xml, and robots.txt contains no "Sitemap:" directive (it also disallows /sitemap/). There is no machine-readable inventory of URLs or per-URL lastmod timestamps; page discovery for this analysis relied entirely on homepage navigation links plus site: search.
Why it matters: Crawlers use sitemap.xml to discover deep pages efficiently and to read lastmod recency signals. Without one, the editorial "Sneak Peak" articles (the freshest, most citable content) and individual show pages may be discovered slowly or incompletely, and no page receives a sitemap-based freshness signal. For a site this large, the absence is a meaningful discoverability gap.
Recommended fix: Publish an XML sitemap (or sitemap index) covering commercial and editorial pages — shows, movies, collections, sports, plan pages, and the Sneak Peak articles — with accurate <lastmod> values, ensure it returns HTTP 200 to all user-agents, and declare it with a "Sitemap:" line in robots.txt.
What we found: The server-rendered marketing/landing templates (home, /live-tv/, /account/signup/live-tv/*, the student-offer page, and marketing-template fallbacks) organize the page under short, all-caps slogan headings — "SO MUCH TO STREAM," "ALL OF SHOWTIME," "THE HOME OF UFC," "PICK A PLAN," "WHAT ARE YOU WAITING FOR? SUBSCRIBE NOW." These are stylistic labels rather than descriptive noun phrases, and the primary H1 is a tagline ("A Mountain of Entertainment") rather than a page-specific claim.
Why it matters: AI engines use the heading outline to segment a page into self-contained, citable passages. Slogan headings carry no standalone meaning ("THE HOME OF UFC" is not a claim an engine can quote to answer "what sports are on Paramount+"), so even the pages that DO render lose passage-level extractability. The same template repeats across many URLs, multiplying the effect.
Recommended fix: On the marketing templates, give each page a single descriptive H1 stating its specific value ("Stream Live NFL, UFC and Champions League on Paramount+") and convert slogan section headings to descriptive H2s that state the section's claim, so each section reads as a quotable answer.
The following items could not be assessed through our analysis method (rendered markdown, which does not execute application JavaScript or expose raw HTML). We recommend your engineering team verify these manually before the validation call.
What to check: Our analysis uses a fetch that returns rendered markdown but does not execute the site's application JavaScript the way a full browser (or a JS-executing AI crawler) would. We observed empty shells on catalog, signup, and Help Center pages while marketing templates rendered fully — strongly indicating client-side rendering — but we can't fully characterize which crawlers execute enough JS to recover the content, nor rule out bot-detection/geo factors.
Recommended action: Verify rendering with JavaScript disabled and with a raw-HTML crawler (e.g., Screaming Frog in list mode, or Google's URL Inspection "rendered vs. raw HTML") across a sample of catalog, signup, show-hub, and Help Center URLs. Compare what a non-JS client receives to the full browser view.
What to check: Our method returns rendered markdown, not raw HTML, so JSON-LD/schema.org markup is not visible; schema coverage is null for all 33 inventoried pages. We do not assert its presence or absence.
Recommended action: Verify with a structured-data testing tool (Google Rich Results Test / Schema Markup Validator) on a sample per template. Add the most specific applicable types — TVSeries/Movie/VideoObject on title pages, Offer on plan pages, FAQPage on Help Center answers, and Article with dateModified on Sneak Peak posts.
What to check: Meta description and Open Graph/Twitter card tags live in the raw HTML <head> and are not visible in our rendered output, so meta_description is null and OG tags are unverified across the inventory. This is a tooling limitation, not a confirmed absence.
Recommended action: Verify with a social-preview/inspection tool or view-source on a sample. Ensure each commercial and editorial page has a unique, benefit-specific meta description and complete OG tags (title, description, image, url).
Partial sample Coverage of some signals is incomplete. Freshness was scorable on only 5 of 33 pages (28 unscored — 22 product/plan and 6 structural pages carry no detectable date), schema coverage was null across all 33 pages (a tooling limitation, not confirmed absence), and with no sitemap, page discovery relied on navigation links plus site: search. Treat these metrics as a directional partial sample and confirm with the manual verification checklist above.
Why now GEO visibility is a timing advantage, and the window is open:
• AI search adoption is accelerating — Akamai measured 1.6 billion daily AI bot requests across its global CDN, up 33% in a single month and 78% over six months (Akamai, February 2026).
• Early citations compound: in narrow consumer categories, the top brands appeared in 70–90% of repeated AI recommendation runs even though the exact list is less than 1% reproducible (SparkToro & Gumshoe, January 2026) — the streaming consideration set concentrates around whoever AI learns to trust first.
• Click behavior is shifting: Google users who encounter an AI summary click a traditional result in just 8% of visits, versus 15% without one (Pew Research Center, July 2025) — being the cited answer matters more than being a blue link.
• Streaming is still early-innings in GEO optimization — acting now means competing against inaction, not against entrenched strategies.
The full audit measures how AI assistants answer the buyer questions your subscribers actually ask — "best streaming service for live NFL without cable," "is Paramount+ worth it now that they raised the price," "where can I stream Tulsa King," "Paramount+ vs Peacock for sports." You'll see exactly which of these return answers that include Netflix, Peacock, Max, or Disney+ but not Paramount+ — and what it would take to appear in them. Because the Layer 1 fixes above restore your own pages to the crawlers first, you'll be improving the baseline before we even measure it.
A 45–60 minute working session where we walk through this document together, confirm the competitive tiers and subscriber archetypes, and resolve the open questions in the Pre-Call Checklist.
We build the buyer query set from the validated inputs and run it across the selected AI platforms, capturing where Paramount+ is cited, where competitors appear instead, and how answers vary by platform.
A visibility analysis, competitive positioning by query cluster, and a three-layer action plan — including the content recommendations that this Foundation Review deliberately holds back until we know which gaps actually cost you citations.
Start now — no need to wait for the call Three technical fixes your engineering team can begin immediately: (1) reproduce and fix the HTTP 500 on the show-hub routes (/shows/tulsa-king/, /shows/star_trek_animated/, /shows/uefa-champions-league/) — a 1–3 day fix that restores your marquee titles to crawlers; (2) publish an XML sitemap and declare it in robots.txt — 1–3 days, restoring deep-page discovery and freshness signals; (3) begin server-side rendering / pre-rendering of the Help Center and catalog pages so support answers, library breadth, and pricing exist in the initial HTML. Note: robots.txt is already confirmed open to all AI crawlers, so there's no access block to clear — rendering is the lever. These don't depend on the rest of the audit and will improve your baseline visibility before we even measure it.
Two jobs before we meet. The questions on the left require your judgment — no one knows your business better than you. The engineering tasks on the right don't require the call at all.