Authexis · SEO

Overview

What this product is really for

A content production platform that interviews you, drafts in your voice, and publishes to every channel -- so consultants, founders, and marketing teams can show up consistently without the ghostwriter bill.

Used as the OG-description template. Tagline shorthand for the homepage: "Publish consistently, in your voice, without the grind."

Domain

authexis.app -- verified in Google Search Console 2026-04-27.

Phase 1 inheritance -- already shipped

Authexis is not starting from scratch. SEO Phase 1 landed 2026-04-21..25:

#2244 sitemap.xml shipped
#2245 robots.txt shipped
#2246 canonical URLs across the site
#2254 first /resources/ long-form article shipped (the content engine works end-to-end)
#2258 .vercelignore one-char fix that was blocking the resources route from deploying
The positioning rule ("AI-assisted is product truth; landing page leads with outcome") is locked in DECISIONS.md and survives this rewrite -- it governs voice on every page below.

What this file adds: the keyword targets, the comparison-page commitments, the cross-link contract, and the SERP-tracking config.

Keyword targets

21 keywords total across three funnel stages.

Bottom-of-funnel (purchase intent)

Keyword	Volume	Difficulty	Rationale
`jasper alternative`	medium	high	Largest competitor; users search this when Jasper output disappoints. Interview pipeline is the differentiator.
`copy.ai alternative`	medium	high	Copy.ai pivoted to GTM/sales -- searchers want content creation.
`ghostwriter alternative for linkedin`	low	medium	Direct ICP language; consultants priced out of human ghostwriters.
`pressmaster.ai alternative`	low	low	Closest competitor; cheaper agency tier is our wedge ($249 vs $699).
`bloomberry alternative`	low	low	Voice-cloning competitor; interview-first is the comparison hook.
`ai writing tool that learns my voice`	low	medium	Voice-fidelity is the buy-in moment.
`linkedin ghostwriter pricing`	low	medium	Comparison shopping; we're priced against $1,500-3,000 ghostwriter cost.
`white-label content tool for marketing agency`	low	medium	Consultant ICP, three-client-plus segment.

Mid-funnel (problem-aware)

Keyword	Volume	Difficulty	Rationale
`how to publish on linkedin consistently`	medium	medium	Core problem; weekly cadence is the win condition.
`how to write in your own voice with ai`	medium	medium	Voice-authenticity is the recurring objection in r/ContentMarketing.
`how to train AI to write in your voice`	medium	medium	Lead-gen sheet target for the voice-system article; sends readers to the Brand Voice Analyzer.
`why does AI writing sound generic`	low	medium	Lead-gen sheet target for the generic-output explainer; frames Authexis as the interview-first fix for missing perspective, examples, constraints, and review.
`how to manage content for multiple clients`	low	medium	Lead-gen sheet target for the marketing-consultant multi-client operating-system article; sends readers to the Team trial.
`content workflow for marketing consultants`	low	medium	Consultant ICP, multi-client lane.
`best way to draft thought leadership posts`	low	medium	HBR's "Has AI Ended Thought Leadership?" debate makes this hot.
`ai that interviews you to write content`	low	low	Almost no competitor uses this exact frame; defensible category language.
`linkedin algorithm authenticity update 2026`	low	medium	News-attached query; ranks fast on a topical post.

Top-of-funnel (awareness)

Keyword	Volume	Difficulty	Rationale
`ai content production platform`	low	medium	Category-defining; thin volume, high intent.
`voice-authentic ai writing`	low	low	New category language we can shape.
`thought leadership content strategy`	high	high	Broad; only worth pursuing with a serious pillar piece.
`ai content marketing`	high	very high	Volume play, low ROI; long pillar with internal links only.

Target pages

13 target pages total.

Slug	Working title	Target keywords	Positioning	Status
`/`	Homepage	`ai content production platform`, `voice-authentic ai writing`	"Publish consistently, in your voice, without the grind."	drafted (Phase 1, may need keyword pass)
`/jasper-alternative`	Authexis vs Jasper	`jasper alternative`, `ai writing tool that learns my voice`	Side-by-side: interview pipeline vs templates.	shipped 2026-05-09 (#2330)
`/copy-ai-alternative`	Authexis vs Copy.ai	`copy.ai alternative`, `ai content production platform`	The gap Copy.ai's GTM pivot left open.	shipped 2026-05-09 (#2331)
`/for-consultants`	For marketing consultants	`content workflow for marketing consultants`, `white-label content tool for marketing agency`, `ghostwriter alternative for linkedin`	Multi-workspace, copywriter-replacement pricing.	shipped 2026-05-17 (#2344)
`/for-thought-leaders`	For thought leaders	`how to publish on linkedin consistently`, `how to write in your own voice with ai`	Solo founder/exec ICP -- voice profile, weekly cadence.	shipped 2026-05-18 (#2345)
`/voice-capture`	Voice capture	`ai that interviews you to write content`, `voice-authentic ai writing`	Interview-first pipeline as the core mechanic.	planned
`/blog/linkedin-authenticity-update-2026`	LinkedIn's authenticity update -- what AI-drafted means now	`linkedin algorithm authenticity update 2026`, `how to write in your own voice with ai`	Topical March 2026 algo piece.	planned
`/blog/replacing-your-ghostwriter`	Replacing your ghostwriter	`ghostwriter alternative for linkedin`, `linkedin ghostwriter pricing`	Voice match + economics.	planned
`/blog/the-interview-first-pipeline`	The interview-first pipeline	`ai that interviews you to write content`, `voice-authentic ai writing`	Why interview-first beats prompt-first.	planned
`/resources/how-to-train-ai-to-write-in-your-voice`	How to train AI to write in your voice, not your vibe	`how to train AI to write in your voice`	Voice system, not prompt tricks; CTA to Brand Voice Analyzer.	planned (#2263)
`/resources/why-does-ai-writing-sound-generic`	Why does AI writing sound generic? It has nothing to work with	`why does AI writing sound generic`	Missing perspective, examples, constraints, and review; CTA to Brand Voice Analyzer.	planned (#2264)
`/resources/how-to-manage-content-for-multiple-clients`	How to manage content for multiple clients without flattening their voices	`how to manage content for multiple clients`	Multi-client operating system: one workspace and voice profile per client; CTA to Team trial.	planned (#2265)
`/pricing`	Pricing	`linkedin ghostwriter pricing`, `white-label content tool for marketing agency`	Anchored against copywriter cost, not SaaS comp.	drafted (needs keyword pass)

Free tools

Top-of-funnel acquisition surface. Each free tool is its own SEO target -- separate keyword shortlist, target page, on-page baseline. Lives at a subpath of authexis.app. Conversion target: trial of Authexis.

Voice analyzer

URL: authexis.app/tools/voice-analyzer
What it does: paste a sample of your existing writing, get an analysis of your voice characteristics -- register, sentence rhythm, recurring rhetorical moves, vocabulary patterns. Demonstrates Authexis's voice-capture engine on a single sample rather than a full interview.
Conversion path: "this is your voice as captured" → "want Authexis to draft in it?" → Authexis trial (start an interview-first profile).
Keyword targets (TBD -- needs research pass): candidates include voice analyzer, writing voice analysis tool, writing style checker, tone of voice analyzer, free writing style analyzer. Volume + difficulty estimates pending.
Status: shell, extraction, result UI, and trial handoff shipped; keyword research still pending.

Named-competitor comparison pages

Confirmed 2026-04-27. Comparison pages are part of the plan, not skipped.

vs Competitor	Rationale	Priority	Status
Jasper	Largest brand; users searching this are Jasper-disappointed and ready to switch.	P0	shipped 2026-05-09 (#2330)
Copy.ai	Copy.ai pivoted away from content creation; live churn intent.	P0	shipped 2026-05-09 (#2331)
Pressmaster.ai	Closest direct competitor; pricing wedge ($249 vs $699).	P1	shipped 2026-05-17 (#2342)
Bloomberry	Voice-cloning peer; interview-first is the comparison hook.	P2	shipped 2026-05-17 (#2343)

Cross-link contract

Inherited from fleet SEO plan.

Footer link to paulwelty.com/tools on every page of authexis.app -- text: "Other tools by Paul Welty" (no nofollow, no marketing-speak). Shipped 2026-05-09: <PaulWeltyToolsLink> component in web/components/cross-link.tsx, used in app-shell footer (every logged-in page) + every public footer (homepage, pricing, blog, blog/[slug], privacy, terms, comparison pages). Component test pins the exact text + URL + rel so future copy refreshes can't silently break the contract.
Reciprocal: a card on paulwelty.com/tools linking to authexis.app.

Links

First-party links we control between Synaxis surfaces. Third-party backlinks (the organic kind) live in Search Console; not enumerated here.

Inbound (other Synaxis-controlled surfaces linking TO authexis.app)

From	Anchor text	To page	Status	Notes
paulwelty.com/tools	Authexis	`/`	planned	After Tools page ships (paulos #733)
polymathic-h blog: (post-slug)	(varies)	(varies)	unknown	Audit pending (paulos #734)

Outbound (this product's site linking TO other Synaxis surfaces)

To	Anchor text	From page	Status	Notes
paulwelty.com/tools	"Other tools by Paul Welty"	footer (sitewide)	planned	Cross-link standard, fleet plan

SERP tracking

Tool: Google Search Console API via paulos seo serp-check (paulos #732)
GSC property: authexis.app -- verified 2026-04-27 by Paul

Config block for `paulos.project.toml`

[seo]
domain = "authexis.app"
tracked = [
  { keyword = "ai content production platform", page = "/" },
  { keyword = "voice-authentic ai writing", page = "/" },
  { keyword = "jasper alternative", page = "/jasper-alternative" },
  { keyword = "copy.ai alternative", page = "/copy-ai-alternative" },
  { keyword = "ghostwriter alternative for linkedin", page = "/blog/replacing-your-ghostwriter" },
  { keyword = "ai writing tool that learns my voice", page = "/voice-capture" },
  { keyword = "ai that interviews you to write content", page = "/voice-capture" },
  { keyword = "how to publish on linkedin consistently", page = "/for-thought-leaders" },
  { keyword = "how to write in your own voice with ai", page = "/for-thought-leaders" },
  { keyword = "how to train AI to write in your voice", page = "/resources/how-to-train-ai-to-write-in-your-voice" },
  { keyword = "why does AI writing sound generic", page = "/resources/why-does-ai-writing-sound-generic" },
  { keyword = "how to manage content for multiple clients", page = "/resources/how-to-manage-content-for-multiple-clients" },
  { keyword = "content workflow for marketing consultants", page = "/for-consultants" },
  { keyword = "white-label content tool for marketing agency", page = "/for-consultants" },
  { keyword = "linkedin ghostwriter pricing", page = "/pricing" },
  { keyword = "linkedin algorithm authenticity update 2026", page = "/blog/linkedin-authenticity-update-2026" },
]

On-page SEO checklist (per page)

<title> ≤ 60 chars, primary keyword at front
<meta name="description"> 150-160 chars, sourced from "what this is really for" + page-specific framing
Primary keyword present in <title>, <h1>, meta description, first paragraph, at least one section heading, and naturally repeated 2-4× in the body (no stuffing -- the page should read as the page you'd write without SEO, just with deliberate word choice). Semantic variants alone are not sufficient; the exact target query has to appear.
Open Graph tags: og:title, og:description, og:type, og:image, og:url
Canonical URL (Phase 1 #2246 -- already enforced site-wide)
Structured data: appropriate JSON-LD (Article / SoftwareApplication / Organization / Product)
Sitemap.xml entry (Phase 1 #2244 -- auto-generated)
Robots.txt allows crawling (Phase 1 #2245 -- done)

Authoring discipline for syndicated lead magnets

Per Eli (eclectis 2026-04-28): in a polymathic-h-canonical world, the SEO discipline has to be applied at the source, not patched at the syndication endpoint. Search engines credit the canonical URL on polymathic-h; whatever polymathic-h ships is what authexis ranks for via the syndication pipeline. So the keyword-placement checklist isn't an authexis post-deploy step -- it's a write-time acceptance criterion on every post tagged for syndication.

Acceptance criteria for any blog post intended to syndicate to /resources

Target keyword present in title (or paraphrase containing the exact phrase)
Target keyword present in H1
Target keyword present in meta description
Target keyword present in the first paragraph of body
Target keyword present in at least one H2
Target keyword naturally repeated 2-4× in the body
Slug follows the topic-only-slug rule (no date prefix; see DECISIONS.md)
Related semantic phrases woven in -- keyword density without keyword stuffing

Failure mode this prevents: target keyword captured in frontmatter or sheet only, body uses semantic variants -- captured-but-unused. Lossy by the next quarter, useful at write-time only. Caught on #2254 the first time a post shipped with the keyword in frontmatter and slug but missing from title and body.

The /humanize pass remains required after the keyword pass. Keyword inclusion never overrides voice; the choice is between natural inclusion and natural exclusion, not between good writing and SEO. If the phrase doesn't fit naturally, the keyword target is wrong, not the writing.

Cross-product overlap (relevant to Authexis)

From the fleet keyword research overlap section:

ai content: Authexis owns "ai content production / writing / drafting." Eclectis owns "ai content curation / scoring / briefing." Never compete on the unqualified phrase.
thought leadership: Authexis owns "thought leadership content / writing / strategy." Eclectis stays clear; if it ever wants the lane, it does so as "intelligence for thought leaders" -- modifier discipline.
content workflow: Authexis owns "content workflow for consultants / marketing / publishing." Textorium owns "content workflow for hugo / static sites / SSG." Never compete on the unqualified phrase.
ai for writers: Authexis owns "ai for writers / writing." Textorium uses "for prolific authors / bloggers" and never says "ai writing tool."
agency: Authexis owns "content agency / marketing consultant." Diktura uses "saas team" and stays out.
newsletter: Three-way conflict resolved 2026-04-27: Eclectis claims the "newsletter platform" search term. Authexis defers -- the platform is a broader content-production / publishing pipeline, not specifically newsletter-shaped. Authexis does not target generic newsletter platform / tool queries.

Decisions

Locked 2026-04-27.

Sequencing -- comparison-first: Across the entire fleet, the lead investment is named-competitor comparison content (Jasper, Copy.ai, Pressmaster.ai, Bloomberry for this product). Pillar pages still happen, but comparison pages ship first to capture live churn intent (Copy.ai's GTM pivot, Jasper disappointment).
Newsletter platform term -- defer to Eclectis: Authexis is not the "newsletter platform" -- it's a content production platform with newsletter integrations. Eclectis owns that term.

Status log

2026-05-18

/for-thought-leaders pillar page shipped (#2345). Second ICP-targeted pillar; companion to /for-consultants. Primary keyword how to publish on linkedin consistently carried in title (39 chars; 50 with -- Authexis template), description (153 chars), H1, hero intro, and a dedicated cadence H2 ("How to publish on LinkedIn consistently -- even when the week explodes."). Secondary keyword how to write in your own voice with ai gets its own H2 plus a four-card section explaining what a fifteen-minute interview catches that a brand-voice doc can't (named examples, opinions with stake, the writer's own sentence rhythm, words they'd actually say). Slug eyebrow is "For founders, executives, and writers" -- concrete ICP language; the page avoids the "thought leaders" phrase in body copy per CLAUDE.md voice discipline. Trial CTA carries ?ref=for-thought-leaders. No inline JSON-LD (matches the shipped comparison-page//for-consultants convention; structured-data signal stays on buildMetadata's canonical/OG output). Page tests at web/app/for-thought-leaders/__tests__/page.test.tsx pin H1 keyword presence, single-H1 rule, ?ref=for-thought-leaders on every signup link, title/description length envelopes, canonical URL, presence of the secondary-keyword H2, ≥2 occurrences of the primary keyword (title + H2), and ICP language. SERP-tracking rows for both keywords were already wired in paulos.project.toml (#2340). Sitemap.xml route is still missing fleet-wide (carried forward from #2344); tracking separately. All five named SEO target pages slated for ship in this campaign (/jasper-alternative, /copy-ai-alternative, /pressmaster-ai-alternative, /bloomberry-alternative, /for-consultants, /for-thought-leaders) are now live.

2026-05-17

/for-consultants pillar page shipped (#2344). First ICP-targeted pillar in the SEO.md target-pages list; primary keyword content workflow for marketing consultants carried in title (42 chars; 53 with -- Authexis template), description (152 chars), H1, hero intro, and dedicated H2. Secondary keywords (white-label content tool for marketing agency, ghostwriter alternative for linkedin) each get their own H2 section so structured-search signal aligns with the keyword target. Page is a custom layout (not the ComparisonPage shared component, which is shaped for vs-competitor pages) but reuses the homepage/pricing visual conventions -- paper-grain background, gold-ribbon, brand-800 alt section, PaulWeltyToolsLink in footer. Trial CTA carries ?ref=for-consultants. Page tests at web/app/for-consultants/__tests__/page.test.tsx pin H1 keyword presence, single-H1 rule, ?ref=for-consultants on every signup link, title/description length envelopes, canonical URL, and presence of both secondary-keyword sections. No inline JSON-LD on this page (matches shipped comparison-page convention; the pre-write security hook also blocks the Next.js-documented inline-script pattern, so structured-data signal stays on buildMetadata's canonical/OG output for now). No sitemap.ts in the repo yet -- sitemap.xml return is 404 in prod; the route was never landed beyond Phase 1's #2244 claim. Tracking separately.

/bloomberry-alternative shipped (#2343). P2 named-competitor comparison page; final entry in the named-comparison shortlist. Frame is mechanism, not pricing -- interview-first capture (voice + POV from speech) vs Bloomberry's style-cloning from written samples. Side-by-side covers voice capture, what gets captured, primary user, workflow shape, best-fit-when, output editability, repurposing, and pricing anchor. "Choose Bloomberry if..." section concedes the verbatim-corpus and scale-existing-voice cases. Trial CTA carries ?ref=bloomberry-alternative. Title 54 chars; description 154 chars; canonical via buildMetadata. Page tests at web/app/bloomberry-alternative/__tests__/page.test.tsx. SERP tracking row added to paulos.project.toml for bloomberry alternative → /bloomberry-alternative; the earlier pressmaster row was missed in #2342 and is added in the same commit. All four named-competitor pages (Jasper, Copy.ai, Pressmaster.ai, Bloomberry) are now live.

/pressmaster-ai-alternative shipped (#2342). P1 named-competitor comparison page; mirrors the #2330/#2331 ComparisonPage convention. Frame is the pricing + workflow wedge ($99/mo Solo vs Pressmaster.ai's ~$699/mo agency-shaped tier). Hero, side-by-side table (voice fidelity, workflow shape, pricing anchor, output editability), "Choose Pressmaster.ai if..." honesty section, and trial CTA with ?ref=pressmaster-ai-alternative. Title 61 chars; description 153 chars; canonical via buildMetadata. Page tests at web/app/pressmaster-ai-alternative/__tests__/page.test.tsx. SERP-tracking config and on-page baseline already wired; Bloomberry (P2) remains the only outstanding named comparison.

Homepage on-page metadata baseline pass (#2341). Title set to "Voice-authentic AI content production platform" (renders as 57 chars under the layout template); meta description rewritten to 152 chars with both target keywords (ai content production platform + voice-authentic); intro paragraph weaves the primary keyword naturally; SoftwareApplication JSON-LD added to the homepage. OG/Twitter tags inherit from buildMetadata. Phase 1 canonical/sitemap/robots untouched.

SERP tracking config landed in paulos.project.toml -- [seo] block with domain = "authexis.app" and 16 tracked keyword/page pairs (#2340). Authexis is now wired into the fleet paulos seo serp-check command.

2026-05-09

/jasper-alternative shipped (#2330) and /copy-ai-alternative shipped (#2331). Both P0 named-competitor comparison pages live.

2026-04-27 (afternoon)

Decisions locked: comparison-first sequencing across fleet, Eclectis claims newsletter-platform term, Textorium TUI excluded from textorium.app scope. Open questions section closed.

2026-04-27

SEO.md replaced with fleet-aligned structure. Keyword research drafted (fleet keywords doc). SERP tracking on GSC API queued (paulos #732). Named comparisons confirmed (Jasper, Copy.ai, Pressmaster.ai, Bloomberry). GSC property verified. Newsletter ownership and sequencing forks open pending Paul.

2026-04-25

Phase 1 SEO foundation shipped -- sitemap (#2244), robots (#2245), canonicals (#2246), first /resources article (#2254), .vercelignore fix (#2258).