AI Content Writer
1500–2500 word articles with FAQ, schema, and internal-link suggestions.
What it is
Powered by Claude Sonnet, the Content Writer generates full SEO articles around any keyword: H1 + introduction, structured H2/H3 outline, body content, FAQ block, JSON-LD schema for rich results, and suggested internal-link anchors from your own site. Auto-detects keyword language (Arabic/French/Spanish/German/etc.) AND auto-renders RTL for Arabic. <strong>Refund-on-fail across the board</strong>: if Sonnet returns garbage / banned-brand phrases / under 500 words, you don't get charged.
Why it matters
Content is the #1 ranking factor. The bottleneck for most sites isn't ideas — it's the time to write. AI Writer collapses a 4-hour writing session into 3 minutes, then you spend 20 minutes editing for your voice. You'll publish 5-10× more.
How to use it (step-by-step)
Follow these in order. Each step is 30 seconds to 2 minutes.
Step 1
(Optional) Train your brand voice first
On the 'Brand voice' tab, paste 2–5 paragraphs of your existing writing and click Train (1 credit, refund-on-fail). From then on every article sounds like YOU, not generic AI.
Step 2
Pick a target keyword
On the 'Write' tab, type the keyword you want to rank for — ideally a 'Golden' keyword or a quick-win from Keyword Research.
Step 3
Watch the 'Since last article' strip
After the second article, a green strip at the top shows how many new articles + how many you published in the last 14 days — same convention as Site Audit / KR / IL / Competitor.
Step 4
(Optional) Get the SERP brief first
Click 'Get SERP brief' (1 credit, refund-on-fail) to scrape Google → DuckDuckGo → Bing top-10 and see the intent, angle, subtopics (H2s), and required entities you must cover to compete.
Step 5
Choose a word count + country, then Generate (30 credits, refund-on-fail)
Pick ~1,200 / ~1,800 (recommended) / ~2,500 words. We auto-analyze the top 10, write to cover them, and SCORE the article's actual SERP coverage (replaces the old hardcoded 80). If we ever return under 500 words or contain banned-brand phrases (ChatGPT / GPT-4 / OpenAI / etc.), you don't get charged. 30–90 seconds.
Step 6
Check the entity-coverage scorecard
If you ran a SERP brief, click 'Score now' (free) to see how many required entities your article covers vs misses — then patch the missing ones inline.
Step 7
Try a duplicate keyword? We'll ask first
If you already have an article on the same keyword in this project, we surface a confirm panel BEFORE charging the 30 credits — saves you from accidental re-generations.
Step 8
Export the report
The hero top-right has Download · Print · Share — same as Site Audit / IL / KR / Competitor. Branded HTML with KPI mix + status mix + word-count distribution + language mix + top articles + brand-voice profile.
Step 9
Want it on autopilot?
Once you have 2 articles, a green Scribe banner appears: the Content Engine agent that drafts 2-3 articles/week into an approval queue. $19/mo add-on. The Content Writer module focuses on one-shot articles; Scribe handles the cadence.
Step 10
Watch the brand-voice drift banner
After you publish ~10 articles in a language, a violet panel surfaces: 'Your brand voice has drifted — time to retrain.' Click it to review/retrain. The profile is now stored per-language, so you can train EN, FR, AR voices on the same project.
Step 11
Internal links are auto-resolved
When Sonnet suggests an internal link to an existing article, we now resolve the placeholder to the real URL via fuzzy title matching against your existing articles + Internal Links crawl. No more dead '#' anchors.
SERP-derived brief before a single word is written.
Outerank reads the top 10 ranking pages for your target keyword and extracts the topics they all cover, the questions they all answer, the H2/H3 structure that wins. The brief is the ranking blueprint — your article inherits it.
- Real top-10 SERP scrape (no fabricated outlines)
- Topic coverage frequency across rankers
- H2/H3 patterns the winners use
- FAQ block derived from People Also Ask
SERP brief · top 10
Full article, FAQ, schema — done.
One generation produces the complete article: title, H1, body, FAQ block, JSON-LD Article + FAQPage schema. Word target is yours to set; tone matches your project's brand voice (if set).
- Title + meta + body + FAQ in one pass
- JSON-LD Article + FAQPage included
- Word target you control (800–3000+)
- Brand voice match (if your project has one set)
Generated article · excerpt
Title · The Practical Guide to Leather Conditioning at Home
Meta · Restore dry leather in 6 steps — no expensive shop visit.
Schema · Article + FAQPage JSON-LD attached.
Coverage score against the live SERP.
Every article gets scored 0–100 against the top-ranking competitors on the same keyword. The score tells you, before you publish, whether the article is more comprehensive than what's currently ranking.
- Score 0–100 vs. the live top-10
- Coverage gaps highlighted inline
- Readability + word count + structure
- Re-score after edits (free)
Coverage score · vs top 10
Approve, publish, or queue for Scribe.
Approve the draft → saved to your library. Send to Scribe → enters the auto-publish schedule. Or download as HTML / Markdown and ship to your CMS manually. Every approved article is searchable by keyword.
- Library searchable by keyword + status
- Download HTML or Markdown
- Auto-publish via Scribe (Mon + Thu cadence)
- Tracker auto-adds the keyword on publish
After approval
Why Outerank
Things Jasper and Surfer don't connect end-to-end.
The article isn't the deliverable — the keyword going live and ranking is. Outerank closes that loop.
SERP-derived briefs, not vibes
Real top-10 scrape, topic coverage frequency, H2 patterns. The brief is the blueprint of what's already winning.
Coverage score vs live SERP
Before you publish, see if your draft is more comprehensive than the current top-10. No more publishing into the void.
Article + FAQ + schema in one pass
Title, body, FAQ block, JSON-LD Article + FAQPage. Generation outputs a fully shipable artifact, not a starting point.
Scribe handoff built-in
Approve → into Scribe's auto-publish queue. The agent picks up the cadence from there. No copy-paste.
Brand voice match
If your project has a tone set (or learned from your existing content), generations match it. Not generic AI-slop voice.
Auto-tracker on publish
Every article that goes live auto-adds its target keyword to Rank Tracker. Performance loop closes automatically.
What costs credits
No surprises. Browsing and viewing saved results is always free.
Pro tips
- →Always edit before publishing — Sonnet needs your voice and product specifics layered in. Brand-voice training gets you 80% of the way.
- →Run the SERP brief before generating for competitive keywords — it tells the writer exactly which subtopics and entities to include, so you outrank deeper pages.
- →Use the FAQ block in the output — it can earn you a 'People Also Ask' slot in Google, which doubles clicks.
- →The included JSON-LD schema makes you eligible for rich snippets (star ratings, FAQ accordions in results) — keep it when you paste.
- →Every action that costs credits is refund-on-fail. Try freely; if Sonnet hiccups, you get the credits back automatically.
- →Set up Scribe for your most-active project, then use the manual Writer for one-off articles only — it's the cleanest cadence for an Agency client.
Plain-English glossary
Every bit of jargon on this page, explained.
- SERP brief
- A research summary built from Google's current top 10 results for your keyword — the subtopics and entities you need to cover to compete.
- Search intent
- What the searcher actually wants (to buy, to learn, to compare). Your article must match it to rank.
- Entity
- A specific concept, brand, or term Google expects to see on a page about this topic. Covering them signals depth.
- Entity coverage score
- 0–100 grade of how many of the SERP's required entities your draft includes. Higher = more complete than competitors.
- Readability score
- 0–100 measure of how easy your article is to read. Higher is friendlier for most audiences.
- Brand voice
- Your saved tone profile (formality, sentence length, do/don'ts) that the writer applies so articles sound like you.
- JSON-LD schema
- Hidden structured data included with the article that makes you eligible for rich results like FAQ accordions.
Questions, answered
Site Audit FAQs
Why does an article cost 30 credits?
What happens if generation fails?
Can I publish the article as-is?
Do I have to run the SERP brief?
Does it write in other languages?
What if I accidentally try to write the same keyword twice?
Can I download or share the report?
Can I automate article cadence?
Does the SERP brief actually scrape Google?
Why does the brand voice prompt me to retrain?
What's the rolling 30-day quota?
How we compare
Honest comparison — where we win, where the big tools win.
vs. Jasper / Copy.ai
Those tools write short marketing copy. We write full SEO articles with structure, FAQ, schema, live SERP analysis — built for rankings, not just words. Plus refund-on-fail on every paid endpoint, multi-language auto-detect, and a shareable branded report.
vs. Surfer SEO
Surfer is better at in-editor optimization scoring; we bundle SERP brief + entity coverage scoring AND first-draft generation AND brand-voice in one step. Cheaper too.
vs. ChatGPT
Raw ChatGPT writes generic content; we feed it your brand voice + the live SERP + required entities + ban its own model name from the output, so the article is tailored, rank-ready, and doesn't leak 'as an AI language model' phrases.
Ready to try AI Content Writer?
Open it on one of your projects.
Come back here whenever you need a refresher.