Read the SERP before you write a word.
Google’s first page already shows what winning your keyword takes: the intent behind the search, the format that ranks, the depth, the topics, the gaps nobody covers. SERP research reads all of it in about a minute, and hands you the structure to beat it.
Ranking is decided before the writing starts.
Every keyword already has a page one: ten winners Google chose, in public, for anyone to read. What they cover, what format they take, how deep they go: that isn’t a mystery. It’s a published answer key. The old way is to write first and check the SERP later. The answer key sits there unread, and the content ships without a real chance to rank.
SERP research reads it for you. And the read is two jobs, not one. First the baseline: what ranking takes today (the intent to match, the format to build, the depth to hit). Then the part that wins: the tools to beat that baseline (the gaps nobody covers, the expected topics a winner skipped, the questions still unanswered). Not to copy what ranks. To out-write it.
The rest of this page is that read, walked step by step. A keyword goes in; a structure built to win comes out.
Name the keyword. That’s the whole setup.
One field: the keyword you want to win. No configuration, no crawl setup, no waiting on an index. The research agents go straight to the live SERP and start reading.
First, it reads the whole field.
All search results, grabbed in real time. This is the page one Google is serving your audience at the moment you run it, SERP features like People Also Ask and related searches included.
And it doesn’t skim titles. The agents pull the full content of every ranking page and read it end to end. That read is the fuel for every insight the sections below chart.
Why they search decides what ranks.
Every search is someone trying to do something: learn it, compare it, buy it, make it. Google ranks whatever serves that job, so the intent mix on page one is the first call that matters: match it, or the best writing in the world stays on page ten.
The winning format is already on the board.
Guides, tools, videos, product pages: page one shows exactly which formats Google rewards for this keyword, in what proportion. Build the wrong one and you’re not competing, you’re miscategorized.
The evidence reads. You decide.
Confirm the intent and format to target: the SERP’s own pick comes pre-selected, or overrule it. It’s the same call you’d make in the product, so make it here: the rest of the page follows your pick.
Then it reads deeper, only where it matters.
With the intent and format confirmed, the analysis narrows to the results that match them. The rest of page one is noise for your purposes: set aside, not averaged in.
Only results matching your intent and format feed the analysis. The rest never skew a number.
Only results matching your intent and format feed the analysis. The rest never skew a number.
A structure built to win. Then you make it yours.
The evidence assembles into a v1: a structure page one already rewards. It’s a starting point on purpose: you bring the know-how no SERP contains, and the tools below are live. Claim a gap, add a missing topic, trim what the SERP ignores, and watch the outline react.
Every signal on page one. One research run.
Twenty-seven tools for long-form content and commercial pages alike, from the first search to the brief a writer holds, so every piece starts from what Google already rewards for your keyword, not from a guess.
The read
Live page one
Google’s first page, fetched the moment you run it. The agents read every ranking page end to end, not just its title.
Country & language
Search the way your audience does: pick the country and language of the SERP, or leave both on Auto.
Search intent
Every ranking page classified as informational, commercial, transactional, or navigational: what searchers want from this keyword, and the kind of content Google rewards for it.
Content format
The format mix on page one (blog posts, videos, product pages, forums, reviews), with a warning when your pick is one the SERP doesn’t reward.
Length intelligence
A target length and a number of sections, each the median of the pages that match your pick, split across your outline, with per-section ranges on page plans.
Related keywords
Mined from the ranking pages’ own content, with monthly search volume where available. Add any of them to your structure in a click.
SERP features
Featured snippet, People Also Ask, related searches, the AI Overview with the sources Google cites, and more. Questions and searches add to your structure in a click.
Refresh keeps your edits
When the SERP moves, run the research again: the evidence and the suggestion regenerate, and the outline or page plan you edited is kept.
Research depth
Tune how many core topics and subtopics a research run suggests, and the must-have sections per page variant, account-wide or per project.
Long-form content
Suggested outline
A v1 built from the ranking pages that match your pick. Drag, rename, add, and cut it like a doc, or reset to the recommendation anytime.
Competitor outlines
Every matching ranking page expanded into its heading tree. Lift any heading worth having straight into your outline.
Content gaps
Topics relevant to your keyword that page one and your outline both leave out, each with subtopics ready to claim.
Missing topic
The most important topic the ranking pages share and your outline still lacks, surfaced with its subtopics.
Top opportunity
Up to three highest-impact moves for your outline, ranked with the why attached. Step through and claim them.
Uncommon on the SERP
The inverse signal: the one section in your outline the winners barely touch, flagged so you can trim.
Commercial pages
Suggested page plan
Typed section cards (purpose, key points, length range), each tagged must-have, SERP-earned, or baseline. Add a typed section and it arrives filled in to match.
Page anatomy
Every ranking page drawn as a strip of sections: which appear on how many pages, how long they run, where they sit.
Page plan gaps
Two lists: the sections ranking pages include that your plan is still missing, and the sections none of them have, where your page can stand out.
Top opportunity
Up to three strongest moves for your page plan, drawn from the SERP’s questions, competitor-only sections, and the length picture.
Variant & awareness
The plan is built for the detected page variant (SaaS, e-commerce, B2B, local, solution, comparison) and buyer awareness. Change either to steer new sections and the draft.
Page audit
Paste your live page’s URL and see which sections the SERP expects that it’s missing, and which run thin. Then import it and optimize.
The handoff
Draft from research
Start content with the outline or page plan already in place, or let the agent draft the whole piece from the research, streaming live.
Research-grounded agent
The agent that writes your content reads the research as it drafts and edits: the structure, the gaps, the length targets, even competitors’ full text. Attach it to any piece.
Content brief
The research run as a spec sheet: keyword, intent, format, target length, audience and tone, a related keywords checklist, competitors, and the structure.
Share link
Hand the brief to a writer or a client as a read-only link: no account needed, printable, revocable anytime.
Import & optimize
Bring a published page in by URL, Markdown, or copy & paste, run the research on it inline, then optimize it against the SERP.
SEO Optimization
The SEO Optimization panel scores the draft against the same run: the keyword, the target length, and the related keywords you’ve covered.
You, after the read.
The features are the flower. This is the fireball: what actually changes about your work once every piece starts from the evidence.
You stop guessing what Google wants.
The intent, the format, the depth: read from page one, not assumed. Every call you make is the SERP’s own answer, and you can point at it.
You walk in knowing what winning takes.
Before a word exists you hold the winning structure: the topics page one expects, the gaps it left open, the depth it rewards.
You publish with receipts.
Every heading traces to evidence. A client or an editor asks “why this structure?” The brief answers with the SERP, not with a shrug.
You show up where answers happen.
Evidence-first structure is what search rewards, and the same substance is what AI answer engines quote. One honest read, found twice.
Asked before the first run.
The practical ones: what a run does, credits, freshness, sharing, and how the research reaches the writing.
You give it a keyword. It fetches Google’s first page live and reads every ranking result end to end, then charts the evidence: the intent mix, the content formats that rank, measured lengths, the core topics page one expects, the content gaps nobody covers, and the SERP’s own furniture (featured snippet, People Also Ask, related searches). You confirm the intent and format to target, and it builds a structure and outline optimized to compete.
Both. Pick long-form content and you get an article outline built from the matching guides. Pick a commercial page (a product, service, or landing page) and the same analysis reads the winning pages’ structure instead: what every ranking page ships, what none of them do, and the copy depth that wins. This page walks both paths: flip the pick in the middle of it.
One successful SERP research run uses one credit (charged only when the run succeeds, and returned automatically if it fails). Refreshing an existing run uses one too. Every plan includes a monthly allotment, and research credits are separate from the AI writing allowance: running research never draws down your writing budget.
Fetched live at run time: the same page one your buyers see, not a cached database. When the SERP moves, refresh the research: the evidence and the suggested outline regenerate, and the outline edits you’ve made are preserved, never overwritten.
No. It’s a v1 on purpose. It gives you the structure the evidence supports; you make it yours with your own know-how. The tools keep score while you work: Content gaps show what nobody covers, Missing topic flags what page one expects and your outline still lacks, Top opportunity ranks the highest-impact moves, and Competitor outlines let you take any heading worth having.
Yes. The research becomes a brief (a spec sheet carrying the keyword, target intent and format, target length, and the full outline), shareable through a public read-only link that needs no account and can be revoked anytime.
The findings are a structured artifact, not a PDF. Attach research to any content and every agent run reads it as context (drafting against the outline, claiming the gaps by name), and the SEO Optimization panel scores the draft against the same evidence. Research is the first half of one pipeline; the writing is the other half.
Write the piece page one is missing.
The old way publishes and prays. The new way reads the field first: the intent weighed, the shape known, the gaps marked, the structure yours.