Your AI traffic already told you what to write.
Viz reads the evidence - what AI crawlers fetch, what AI engines ask, what you are starting to rank for - and turns it into reviewed, finished drafts in your WordPress. You approve the topics, you hit Publish, and every piece reports back: crawled, cited, visited, converted.
From crawl evidence to published, measured articles.
Most content tools start from a keyword list. Viz starts from your server: the questions machines are already asking of your site are the demand signal, and the pipeline exists to answer them.
The researcher proposes
A daily researcher (weekly on Free) mines your readability scores, AI crawl activity, and rising search queries, and proposes topics across the funnel - awareness, consideration, decision - each with an angle and a rationale.
You approve
Nothing is written until you say so. Approve the topics worth having; reject the rest and the researcher will not propose them again.
Written, then attacked
An outliner and writer produce the piece; adversarial reviewers try to fail it. Only a draft that survives is delivered - into Posts → Drafts, hero image set, 1,200 to 1,800 words.
You publish
Edit it, schedule it, or publish as is - it is a normal WordPress draft. Publishing is the approval; there is no auto-publish and no separate tracking setup.
Viz proves it
From the moment it goes live, the piece is measured: AI crawls, AI citations, referral visits, and what those visitors go on to buy, book or submit. A fresh report covering the last seven days arrives every morning.
Built to be refused.
The reviewers reject more than they pass. Drafts that do not clear them never reach you and never cost a credit.
Topics for the whole funnel
Every research run proposes topics across awareness, consideration, and decision - labeled, so you can see at a glance whether a piece attracts new readers or closes the ones you have. Stages with nothing worth writing are skipped, not padded.
Comparison articles, with receipts
When your visibility tracking shows a competitor winning prompts you should own, the researcher can propose a comparison piece. Claims about the competitor come from a dossier built off their own public pages, and the reviewer refuses any draft that makes an unsourced competitive claim.
A reviewer that can say no
Drafts are written and then argued over by adversarial reviewers before you ever see them. If no attempt clears the quality bar, no draft is delivered - and the piece is never billed. You only review work that survived review.
Real articles, not filler
Drafts land at 1,200 to 1,800 words with the shape AI answers reward: the question answered early, claims grounded, no throat-clearing. The reviewers explicitly reject thin pieces and stock AI phrasing.
A hero image, already set
Each draft arrives with a generated 16:9 hero illustration, set as the featured image with alt text. Swap or remove it like any featured image - if your theme already set one, Viz leaves it alone.
On-brand by default
Viz reads your live site - fonts, colors, imagery style - into a brand kit, and hero images are generated from it and reviewed for design quality before they ship. Drafts look like your site made them, not a stock template.
Every piece, one board
The pipeline page sorts everything into tabs: Drafts, Published, In progress, Dismissed, Failed. Dismiss a draft and Viz asks why (optional) - your feedback travels with the draft and improves the next ones. Failed runs show what went wrong, with one-click retry.
It learns what works
Each week Viz distills the results - which pieces got crawled, cited, and visited, and why you dismissed the ones you did - into an editorial playbook that steers the next round of topics and drafts.
Proof, every morning
Publishing a draft is the approval - tracking starts on its own. Every published piece is measured for AI crawls, AI citations, and referral traffic, per URL, with a fresh report covering the trailing seven days in your dashboard every morning.
Content that reports to the same dashboard that inspired it.
The crawls that suggested a topic and the crawls that pick up the finished article land in the same analytics. Citations show up in AI Visibility, and the humans who click through show up as referral visits, with the purchases, bookings and leads they go on to make credited to the assistant that sent them in AI Conversions. No exports, no attribution spreadsheet, no Google connection: publish, then watch the piece move through the funnel it was written for, all the way to converted.
AI Content Pipeline FAQ.
Where do the topics come from?
From your own site, not a keyword database. The researcher reads your machine-readability scores, which pages AI crawlers actually fetch, and the search queries you are starting to rise for, then proposes topics with a stated angle and rationale. On paid plans it runs daily; the free plan gets a fresh brief every week.
Does anything publish without me?
No. Drafts land in Posts → Drafts and wait. You edit them like any WordPress draft, and hitting Publish is the approval - that is the moment tracking begins. Dismiss anything you do not want, and tell the pipeline why so the next batch is better.
How do article credits work?
One credit is one delivered draft. Lite includes 10 a month, Pro 50; unused articles do not roll over. Failed runs and reviewer refusals are never billed - if the pipeline cannot produce a draft worth your time, it costs you nothing.
What stops it from producing generic AI filler?
The pipeline is built around rejection. An outliner and writer produce attempts; adversarial reviewers score each one against a rubric that treats thinness, unsourced claims, and stock AI phrasing as failures. A draft below the bar is refused outright rather than delivered, so your Drafts folder only ever contains pieces that survived.
Can it write comparison pages against my competitors?
Yes, carefully. When your AI visibility tracking shows a competitor winning prompts you should own, the researcher can propose a comparison topic. The pipeline then builds a dossier from the competitor's own public pages (fetched politely, robots.txt respected), and the reviewers hard-fail any draft that makes a claim about a competitor without a source behind it. You approve the topic and publish the result like any other draft.
What does the free plan include?
Weekly topic briefs, so you can see what the researcher would have you write, plus one full trial article - approve a topic and watch the pipeline research, write, review, and deliver it. Ongoing generation starts on Lite.
How do I know the articles are working?
Every published piece is enrolled automatically and measured per URL: AI crawls, AI citations, and the referral visits assistants send back. Three tiles at the top of the pipeline page keep the score - pages published, pages crawled by AI, pages receiving AI traffic - and a written report covering the last seven days lands every morning.