A clearer plan.
Content that reads as you.

Vizebel reads how your website describes the business and names the broad spaces it already claims. A subject you happen to have written a lot about doesn’t automatically become one.

Archive the areas you’ve stopped pursuing, and add any the site has never mentioned. What remains on the list is the strategy; there is no hidden ranking or fixed number of slots.

This is where a broad direction becomes work. Topics are drafted from what your audiences need and ask, then placed in the focus area they serve. Pages, keywords and AI answers stay attached to the topic, where they can be acted on.

For each topic, Vizebel gathers what customers search and ask from Search Console, People also ask, Google autocomplete, related searches, YouTube and Reddit, and shows which page answers each one, which pages only partly answer, and what has no page yet.

A flat topic list can’t show this. Some focus areas have no topic behind them: you claim them, but nothing in the plan serves them. It works the other way too, when customers clearly need a topic your positioning never mentions.

Beside the charts on every screen, Viz says what they show in plain words and names the next move, with where to start and a link to it. On Focus areas it names the signal holding your strategy back. Ask a follow-up straight from the note.

Ask Viz to review or rewrite a page, start a draft, gather customer questions, check AI answers or manage audiences. Anything that costs credits or removes something shows as a card with its price, and runs only when you tap it.

Each topic and focus area gets a health score from your pages and your reach. Your pages: do they answer what customers ask, and do they hold up. Your reach: are you found in Google, do you earn the click, do AI answers name you, and does the traffic do anything. A signal only counts once it has been measured. Until then it’s unknown, and it doesn’t drag the score down.

Each topic carries buyer questions to test in AI answers, each with what the buyer is trying to do, from understanding a problem to choosing a shortlist.

You get something concrete to react to on day one. Vizebel reads your website and drafts the audiences it seems to serve. They’re suggestions for you to correct. Nobody surveyed your market.

Each audience records the situation that brings them to you and what they hope to get out of it. The content starts from there.

Note what customers say, and what they would never say. That vocabulary carries into the questions, briefs and drafts.

When several audiences want different things, mark the one who should lead. The others stay in the plan.

You know your customers better than your website does. Correct any field directly, add an audience from a description, your notes, a brief or a call transcript, or redraft one in place from new notes.

Vizebel suggests audiences your site doesn’t speak to yet, for you to add or dismiss.

Audiences don’t disappear after setup. They sit on the topics and questions they care about and steer the angle and language of what you write.

Each audience gets a card read in two halves: your pages (how much of what it searches and asks a page answers, and how many of its topics are covered) and your reach (how often AI answers show you for its questions, and the search clicks its pages earn). Articles that serve no audience are listed underneath.

An article, a product page and a service page are each judged against what that kind of page has to do, so nobody tells you to add a blog introduction to your pricing.

Each page is scored on whether it answers the question it is for, and on whether it is original and trustworthy: first-hand detail, and writing that reads naturally. FAQ blocks, summary boxes and markup earn nothing on their own.

“This page is weak” isn’t much to go on. The review lists the questions it should answer but doesn’t, and the facts, examples or proof that would make it convincing.

A rewrite shouldn’t repeat a mistake. So the review lists the statements a reader would act on that could be wrong, each beside its sentence on your page. Confirm each one, change it or leave it out, and the rewrite follows your answer.

The review asks for up to five first-hand examples the page lacks, and the rewrite puts each answer where it belongs.

When a question is won by a checklist or a comparison and you have an essay, Vizebel redrafts it in that shape from what is already there. Reshaping is for articles; product and service pages get the review and a facts checklist to work from.

The draft sits beside your page and keeps its images and videos. Anything still missing is marked as a gap to answer or leave out, and once none are left you copy it as HTML or Markdown. Your published page stays exactly as it is until you replace it yourself.

Review a finished draft on the same checks as the page, beside the live page’s score. Once you publish, Vizebel reads the page again and says how much of the draft it carries.

Location, service and category pages that differ by a few swapped words are grouped, with the fix for that kind of page, and a “last updated” date that moved without the text changing is flagged.

What actually happened: the searches your pages appear for, the clicks they earn and what those visitors did next. Read-only, free to connect, and synced daily.

How many visitors ChatGPT, Gemini, Perplexity, Copilot, Claude and others send you, by assistant and by the page they land on, from Google Analytics. They’re counted apart from organic visits, and they’re a minimum, since an app that hides where a visit came from shows up as direct.

No second tool needed for the terms you care about. Paste in a list you already have or take suggestions per subject, and see each day’s Search Console position and the page Google is showing. Suggestions name only the places you serve, read from your site and yours to edit.

Vizebel asks your chosen questions of up to four AI engines: ChatGPT, Gemini, Copilot, Claude or Google’s AI Overview. It keeps the whole answer for you to read. Choose which engines a workspace asks, and a run costs only those.

Who gets recommended in your place, question by question and engine by engine. Expect a few surprises. Add a site that AI answers cite to your competitors straight from the topic page.

Buyers rarely stop at one question. Add a qualifier such as team size, budget, location or an integration, and see whether you are still named.

This one can quietly sink everything else: whether search engines and AI crawlers are allowed to read your site at all, from your robots rules to your Cloudflare settings.

Whether the pages you care about are in Google at all, and which ones carry a setting that is keeping them out.

Vizebel checks that your sitemap covers your pages, and whether you offer an llms.txt for assistants.

Pages missing from your sitemap, pages nothing links to, links to pages that fail or redirect, and broken links to other sites. Each finding opens onto the page, the link text and the fix.

Organisation, article, FAQ, how-to and product markup are checked for presence and correct use, and each article for a named author with a byline and profile link.

We check for pages that fail to load or take too long, and for videos with no transcript or chapters for an assistant to read.

Vizebel reads pages the way the web’s standards describe, never tuned to one site. If your site only works one particular way, that shows up as a finding you can fix.

Ignore a finding you have decided to live with, and it stays quiet until it gets worse.

You shouldn’t have to translate this for your developer. Download the findings as a PDF and send it to your developer or website provider.

Related ideas, customer searches, keywords and AI-answer questions become planned articles, each with its audience, its sections, the evidence behind it and why it ranks where it does. Where you already have a page, the plan is to improve it. Each topic keeps three open at a time.

Each idea carries the question it answers, the engine that raised it and the source behind it. Judge it before you commit to anything.

When a rival’s page beats yours, Vizebel reads the pages an engine quoted and sets out what they have that yours doesn’t, from how the page is built to the questions it answers.

Some questions are answered by review sites, directories and roundups, and those ideas are marked as a listing to go after. Writing a page won’t win them.

A brief names the question the piece answers and the format that tends to win it: a guide, a checklist, a comparison, a case study.

A short interview asks what you know, the numbers you can share and how you actually do the work before the outline is written.

A draft started from a planned article carries its searches, keywords and questions as a checklist you can trim, and the draft shows which it covers, mentions or still misses.

Drafts are written in Studio, in your tone of voice, learned from your existing pages. You review the outline, the draft and the final, with the claims to check listed, before publishing in your own tools.

Impressions, clicks and visits from search over time, for the site and for each subject you chose.

Searches for your own name are separated from everything else, so a campaign or a mention doesn’t read as your content working. What’s left shows whether you’re reaching people who hadn’t heard of you.

Each topic charts its health month by month, with the change since last month beside each signal. Small improvements show up.

Every run is kept. Compare how often you were named across runs, question by question. Switch a snapshot to re-run monthly and the comparison keeps itself current.

The month in plain words, with your logo on it. Add your own commentary, or let AI draft it from the figures for you to edit. Send it as a private link, or print it. AI answers are compared with last month on the same questions, so you can see which ones you won and lost.

Every day, weekdays or Mondays, at the time you pick. The Today page keeps a record of what Viz picked each day, what ran and how it ended.

Refreshing your Google data and pages and checking your site are free, so they run without asking.

It says what finished, what’s still going and what’s waiting for you. Approve costed steps from the email or from Today.

Page reviews, snapshot re-runs and question gathering can run without asking if you allow them. Page rewrites and changes to your strategy always wait for you.

Planning a run uses 2 credits. That and anything Autopilot runs on its own stay within the cap you set. If the cap or your credits run out, we email you, and you can run it again from Today once there’s room.

Autopilot works inside Vizebel. You check the work and publish it in your own tools.

MCP is the standard way AI apps connect to other tools. Pick your app in your Vizebel account and copy what it shows: a connector URL, a command or a config entry. Then sign in to Vizebel to approve the connection.

Ask your AI app about your business and it reads your topics, pages, Google data and AI answers straight from your workspace.

It can review and rewrite pages, gather what customers ask, accept planned articles and work through a Studio draft step by step. Drafts land in Vizebel, where you check and publish them.

Anything that costs credits is quoted first and runs only when you agree, within a monthly limit you set for the connection.

Pick the workspaces, and whether it may read, make changes or spend credits. It can’t reach billing, API keys, invitations or organisation settings.

Nothing disappears in bulk. Taking off a keyword, question, topic or focus area is approved one at a time, and a removed topic or focus area is archived, so it can come back.

Each call is listed under the connection for 90 days. Disconnect one connection, or switch MCP off to end them all.