Skip to content

Content Generation

ReachLLM turns your visibility data and brand knowledge base into publish-ready content. This guide covers AI articles, the Strategy execution plan, Reddit drafts, and where you can publish the results.

Every Content Studio project runs through the same shared intake:

  1. Choose the outcome: start new content, improve an existing page, or write a new website page.
  2. Choose the content type: select the format and its specific research, structure, schema, review, image, and release rules.
  3. Research inspiration: use proven web pages, real Brand Visibility citations, competitors, internal pages, uploads, and approved brand knowledge.
  4. Plan and draft: review the outline before generating the full content.
  5. Create assets: use the suggested featured-image prompt, generate another option, or insert a saved brand asset.
  6. Review and approve: teammates and the ReachLLM agent improve the same document before release or handoff.

If live AI research is temporarily unavailable, Content Studio still offers six concise topic suggestions derived from the selected prompt, target queries, industry, and brand context. You can pick one, edit it, or enter your own topic without waiting for the provider to recover.

Improving an existing page is content-only. ReachLLM reads the current copy into source context, keeps accurate facts and the original purpose, researches useful references, and creates an improved document plus suggested assets. It does not rebuild the page design in Content Studio.

Writing a website page is also content-first. Content Studio creates and approves the page content. The Website workspace remains responsible for applying your site's design system, layout, components, and code after an explicit handoff.

Because content is grounded in your knowledge base, keeping your brand profile up to date directly improves output quality. See the Brand section of the app to review and edit your knowledge base.

Content Studio is the shared workspace for blog posts, guides, reviews, comparisons, listicles, technical articles, case studies, About pages, pricing pages, landing pages, product and service pages, documentation, FAQs, newsletters, and social campaigns. Open Content Studio from the Optimize section of the main navigation. From the same Create content menu, you can start new content, improve an existing page, or write a new website page.

Each format is a versioned content profile. A profile defines the sections that belong in the document, what the research should prove, which review checks matter, which schema type is valid, what image suggestions are useful, and where the approved content can go. This keeps the workflow modular so ReachLLM can add or update a format without creating another disconnected generator.

Content Studio confirms the new document before the intake opens, then saves every setup change and the text draft through the same secure workspace connection before image generation begins. You can leave the page while assets are still being prepared and return to the same document from the library. A save failure stays visible, keeps your current work on screen, and stops navigation until you retry.

On phones, the document stays full width while comments, the ReachLLM agent, reviews, internal links, sources, assets, metadata, and publishing controls open from the compact top bar.

Inside a document, you can:

  • edit the same draft with teammates in real time
  • see each active teammate once, with their real profile photo and the passage they are working in
  • select text and press Command-I (or Control-I) to give the exact passage to the ReachLLM agent
  • review an agent edit as an insertion and deletion before accepting it
  • leave anchored comment threads, mention teammates, reply, resolve, and reopen
  • run basic checks for structure, links, metadata, alt text, and readability
  • run deep review agents for brand voice, approved facts, content depth, competitors, SEO, and answer-engine usefulness
  • find internal links from real website, audit, and Content Studio page data, then review the link as a proposed edit
  • inspect the sources and inspiration used for the document, including visibility citations, competitors, uploads, internal pages, and live web research
  • insert saved brand assets, upload a reference, generate an image, or import Figma colors and typography
  • edit search metadata and preview the schema type required by the selected content profile
  • move the document from Draft to In review, Changes requested, Approved, and Published

Agent edits are proposals. They carry the document revision and selected passage they were created from, so an old proposal cannot silently overwrite newer teammate work. Accepting a proposal records a new version. Rejecting it leaves the document unchanged.

If an approved document changes, it returns to Changes requested so the published revision always has an explicit human approval.

The Content Studio agent is the same ReachLLM agent available from the main Agent page, with additional document context. It can use:

  • the active document and selected passage
  • approved brand facts, products, services, and writing style
  • the saved competitor set
  • website pages and existing content
  • Search Console and analytics data when connected
  • AI visibility prompts, answers, citations, and source research
  • uploaded company knowledge and source files

The selected passage appears as a neutral quoted attachment inside the Agent composer and remains visible on the sent message. Remove it before sending if you want the agent to answer about the full document instead.

The embedded Agent uses the same composer and capabilities as the main Agent page. You can attach files, dictate, type @ to mention the active brand or a competitor, use /websearch, turn on Plan mode, and open connected data sources. While it works, the panel shows the ReachLLM mark, plain-English thinking and tool activity, plans, sources, charts, and follow-up actions.

Each document keeps one shared Agent conversation. Teammates who open the same draft can watch the same tool activity, continue the thread, and redirect the Agent without copying chat links. Opening another draft never shows messages from an unrelated Agent chat.

Comments are anchored to a document block and, when available, the selected passage. A thread shows the author's profile photo, replies, and timestamps. When resolved, it records who resolved it and when. Resolve a thread when the decision is complete, or reopen it later.

Live presence is temporary, deduplicated by account, and only shows active collaborators. Opening three tabs does not create three people. Versions and accepted agent proposals remain available as an editorial audit trail.

Owners and admins can approve and publish. Editors can write, comment, and run reviews for the brand projects they can access. The release panel shows the profile of the teammate who approved and the teammate who published, with the revision and time.

The Assets panel uses the same brand library as the Knowledge Base. A generated image can use one or more saved references, a purpose such as article hero or inline diagram, and a chosen aspect ratio. Inserted images retain their hosted URL, alt text, caption, dimensions, prompt provenance, and brand asset ID.

When an image provider or upload fails, Content Studio keeps the saved draft, marks the unfinished image slot, and offers a retry from the Images or Assets panel. It no longer leaves a failed image looking as if it is still generating. If only part of an image plan completes, each successful image stays in its intended slot and the remaining slots stay available for retry.

Figma import adds color and typography tokens to brand context. It does not copy an entire Figma page into the article. Connect Figma in Account settings > Integrations, then paste a Figma file URL in the Assets panel.

Basic review is deterministic and fast. Deep review runs specialist agents in parallel and stores each finding against the relevant block. Findings remain in Open or Resolved so the team can see what changed and why.

The Links panel ranks real internal pages against the current document. It shows the target URL, suggested anchor, reason, and available search signals. A link is created as a reviewable proposal and never inserted silently.

The Sources panel keeps research visible beside the draft. Selected competitor pages now remain attached to outline generation. Brand Visibility citations retain their source URL, title, provider, prompt, excerpt, and retrieval time so the team can inspect the inspiration behind the content.

Before approval or publishing, complete the release checks required by the content profile. The featured-image check reads the same selected image used by preview, schema, and publishing. For formats that require one, the Assets panel starts with a useful brand-aware prompt so you can generate an image without writing a detailed prompt first.

The editor and publishing service enforce the same release contract. Website and CMS formats verify the meta title, URL slug, 120 to 165 character description, exact structured-data type, required structured-data fields, featured-image requirement, and image alt text. Newsletter and social formats skip website-only metadata and schema checks, then download the latest shared document as Markdown.

Approval follows the selected content profile. A how-to guide can use HowTo, an FAQ resource can use FAQPage, a product page can use Product, and other formats keep their configured schema type. ReachLLM does not convert a valid profile-specific schema back to Article or BlogPosting during approval.

An approved website page can be sent to Website from the release panel. The handoff is pinned to the approved revision, so later document changes cannot silently alter the Website draft. Website then applies the existing site design system and owns preview and publishing.

Every article, site page, and page rewrite follows the same SEO and AEO strategy, so the output is built to be cited by AI engines, not just to read well.

  • Audience first, keywords second. Each topic is traced through audience, product, topic, then keyword. If a topic does not connect to real product value, it is not written.
  • Intent over volume. The format matches the buyer's funnel stage: top-of-funnel explainers (what is, why), middle-of-funnel how-to guides and frameworks, and bottom-of-funnel comparison and product pages (best, vs, alternatives, pricing) that convert and win high-intent AI citations.
  • Structured for extraction. AI engines quote passages, not whole pages, so content leads with a direct answer, uses one H1 with a clean heading hierarchy, and includes scannable lists, tables, and a real FAQ section.
  • Citable by design. Claims are backed by authoritative sources and statistics with dates, the tone stays authoritative rather than promotional, and marketing adjectives are avoided because AI engines down-rank them.

For competitor comparison pages (yourbrand vs competitor, competitor alternatives, best category), ReachLLM uses a narrow-concession structure that AI Overviews and answer engines tend to quote:

  1. Concede exactly one narrow, genuine niche to the competitor, then claim the broader ground for your brand.
  2. Lead with a short, quotable verdict block directly under the title: one concession, one broader claim about your brand, and three concrete differentiators.
  3. Follow with an at-a-glance comparison table with sourced rows, honest competitor strengths kept inside that one niche, your broader strengths, a feature-by-feature dive, and clear "choose X if / choose Y if" guidance.
  4. Add a FAQ that answers the exact comparison questions people ask (including "who should use the competitor instead"), plus FAQPage, Product, and BreadcrumbList structured data and a self-canonical URL.

You can ask the ReachLLM agent to build a comparison page directly, for example "build a comparison page for us vs [competitor]", and it will research, write, and create the page in your Website workspace.

AI articles are limited per month by plan:

Plan AI articles per month
Pro 10
Scale 50
Enterprise Agencies 50
Enterprise Custom

On Scale and Enterprise Agencies, limits are pooled across your brand projects. See Plans and pricing for full plan details.

Strategy gives you a focused action center of recommended items, each one a concrete action to improve your AI visibility. The Today tab groups actions by setup, knowledge, visibility, on-page work, content, and off-page work so you can filter the noise and work on the highest-impact items first.

  • Blog posts
  • Social posts
  • Social comments
  • Directory listings
  • PR emails
  • Page rewrites
  • GEO fixes
  • Brand knowledge updates
  • Writing style updates
  • Brand fact review

Each action has a detail panel with rationale, evidence, and the next place to go. Some actions open the Knowledge Base, writing style, Brand Visibility, GEO Audit, PR Outreach, Account integrations, or the agent. Content and social actions still open the Strategy drawer for draft review.

Use Done when you completed an action and Decline when it is not worth doing. Both states are saved to the brand's Strategy plan in Firebase, so completed and declined actions do not come back as active work. You can still use the Calendar, Backlog, and Published tabs to manage content timing and review finished work. Daily caps on generated items depend on your plan tier.

PR outreach actions are filtered to realistic targets. ReachLLM avoids outreach tasks for sources that usually cannot be contacted directly, such as Wikipedia, YouTube, social networks, app stores, review marketplaces, competitor sites, and broad public repositories. When a cited source is useful but not reachable, Strategy should recommend content, directory, or social work instead.

ReachLLM drafts Reddit posts tailored to relevant subreddits. Each draft goes through rule compliance checks against the target subreddit's rules, so you are less likely to post something that gets removed by moderators.

Review every Reddit draft before posting: community norms vary, and a human read is the best final check.

When content is ready, you can publish it to:

  • WordPress: via the ReachLLM Connector plugin or a WordPress application password. ReachLLM uploads the hero image into the WordPress Media Library when the connected user can upload media, then sets it as the featured image. See the WordPress plugin guide.
  • GitHub: content arrives as a pull request you review and merge. Generated article hero images are committed into the repository alongside the article file by default, so the post can reference repo-hosted images instead of only external URLs. See the GitHub integration.
  • Webflow: content can publish into a CMS collection as a draft or published CMS item. ReachLLM uploads the hero image into Webflow assets when the token allows it, then maps that URL into a matching collection image field when the collection exposes one. If asset upload is unavailable, the original generated image URL is still used. See the Webflow integration.
  • Shopify: content can publish as a Shopify blog article draft or published article. ReachLLM uploads the hero image into Shopify theme assets when theme permissions allow it, then sends that URL as the article image. If theme asset upload is unavailable, the original generated image URL is still used. See the Shopify integration.
  • Copy or export: take the content anywhere else.

Articles in Content Studio publish through the same Content & CMS adapter contract as the agent. That means an existing generated article can be sent later to WordPress, GitHub, Webflow, or Shopify with its title, slug, excerpt, body HTML, schema, and hero image metadata preserved where the destination supports those fields. For WordPress, GitHub, Webflow, and Shopify, generated hero images are uploaded into the destination media/assets system when the connected token or user role has the needed permissions.