One-shot AI content
Fluent paragraphs, structure from nowhere — the average of everything the model has ever read, aimed at no one's actual search. It sounds finished. It isn't decided.
FOR SEO AGENCIES · IN-HOUSE TEAMS · SITE OWNERS
Research-backed page blueprints, built to rank.
Intent Atlas reads everything that ranks for your keyword — the top 20 results captured, every page fetched and read — and delivers the complete plan for the page you should build: blueprint, writer's guide, build guide, and a working HTML mockup.
FROM $199/MO · EVERY PLAN IS THE WHOLE PRODUCT · FAILED RUNS RETURN THEIR CREDITS
Newly launched — no borrowed logos, no invented case studies. Judge the sample package and the method; every account is onboarded personally.
RESEARCH TRAVERSE — PREMIUM ON-PAGE
SPECIMEN
best 4 season tents
northwindoutdoors.com · guides/4-season-tents
18pages in evidence
130sections read
12query groups
THE ROUTE EVERY RUN WALKS · WATCH IT LIVE, STATION BY STATION, IN THE APP
Top 20ranking results captured per run, localized to your market
7 stationson every run — each one streamed live in the app
4 deliverablesin every package: blueprint, writer's guide, build guide, mockup
100%of a run's credits returned automatically if it fails
LAUNCH EDITION — NO BORROWED LOGOS, NO INVENTED CASE STUDIES. JUDGE THE ARTIFACT INSTEAD: VIEW A FULL SAMPLE PACKAGE →
THE PROBLEM
Most pages are written from a guess.
Somewhere behind every underperforming page there's a thin brief: a keyword, a word count, a list of terms to include. The structure — what the page actually says, in what order, for whom — got decided by nobody.
Score-chasing checklists
Term coverage isn't intent. A page can mention every phrase on the list and still leave the searcher's actual question standing — a high score on the wrong page.
Hand-built briefs
Read the whole SERP, take every page apart, decide the structure on evidence. Done properly it's a day of senior time per page — so it almost never gets done properly.
The evidence for what a page needs already exists — it's sitting in the pages that currently rank. Reading all of it, properly, is pipeline work. So we built the pipeline.
THE PIPELINE
One keyword in. The whole SERP, surveyed.
Every run walks seven stations — the same seven you watch live in the app. Ranking pages are read as evidence of what searchers need, then the blueprint is argued from that evidence, section by section.
Competitor pages are evidence, not authority. THE RULE EVERY RUN IS BUILT ON
SAMPLE RUN #412 — MID-FLIGHT AT STATION 4
Collect
The top 20 results for your keyword, captured and localized to your market.
Fetch
Every ranking page fetched whole — pages, not snippets or summaries.
Read
Every ranking page read in full — hundreds of pieces of evidence per run.
Map
Queries and evidence grouped into an intent map: what searchers are trying to get done.
Compose
The blueprint composed from evidence — each section with its job and its reasoning.
Guides
Writer's guide and build guide written per section; the HTML mockup assembled.
Package
Final checks, then one package lands. If a run fails, its credits return in full.
STAGE 4/7
Map intent
STARTED 12:41 · ~9 MIN LEFT
46 of 130 evidence sections mapped · 12 query groups traced
18pages in evidence
130sections read
12query groups
Runs are glass-box: every station streams its work in the app while it happens.
ANY KEYWORD, ANY MARKET — SAMPLE RUNS:
best 4 season tents
small business retainer agreement
cast iron skillet care
emergency dentist austin
how to winterize a campervan
THE PACKAGE
Four deliverables. One run.
Every blueprint run — premium or standard, on-page or guest-post — delivers the same four-part package. Plates I–IV.
PLATE I — THE BLUEPRINT
The page, decided.
A section-by-section plan: what the page covers, in what order, and why. Every section carries its job — the question it resolves for the reader — and the evidence that earned it a place. The structure is argued, never assumed from a template.
- Every section: its job, its position, its reasoning
- Reviews your existing page when there is one
PLATE I — INTENT MAP VIEWSPECIMEN
- “just show me the best ones”→ SEC 01
- “what does 4-season actually mean”→ SEC 02
- “tent for winter basecamp”→ SEC 03
- “wind rating / snow load numbers”→ SEC 04
- “will a 3-season tent survive winter”→ SEC 05
12 QUERY GROUPS → 14 SECTIONS · EACH MAPPING ARGUED IN THE BLUEPRINT
PLATE II — THE WRITER'S GUIDE
Hand it to any writer.
Per-section writing guidance: what the section must resolve, what to include, how deep to go, what voice it takes. A writer who has never seen your SERP can execute the page — the research rides along with the instructions.
- Written per section, not per page
- Works with any writer — human or AI
PLATE II — WRITER'S GUIDESPECIMEN
SECTION 04 — THE SPECS THAT MATTER
- Must resolve: “how do I judge any of these tents?”
- Open with the three numbers buyers compare — wind, snow load, venting — then what each means in use
- Evidence: 11 of 18 ranking pages list specs; only 3 explain them. Explain them.
- Depth: ≈ 350 words · plain, gear-literate voice
SECTION 05 — WHERE 3-SEASON HOLDS UP …
SECTION 06 — LIVING IN IT: SPACE & WEIGHT …
PLATE III — THE BUILD GUIDE
From approved draft to published page.
Everything between a finished draft and a live URL: the heading tree, internal links with their anchor text, structured-data blocks, image slots with alt text written. Your builder gets instructions, not interpretations.
- Heading tree, internal links, structured data, image slots
- Assembly order a builder can follow cold
PLATE III — BUILD GUIDESPECIMEN
- H1Best 4 season tents: picks that hold in real weather
- H2The short answer
- H2What “4-season” actually means
- H2Picks by use
- Internal links — 6 planned, anchor text written
- Structured data — 2 blocks, ready to paste
- Images — 5 slots, alt text written
PLATE IV — THE MOCKUP
See the page before you write it.
The blueprint rendered as a working HTML page — real structure, real hierarchy, ready to walk through. Click it, share it, put it in front of a client or a developer before a word of copy is drafted.
- Working HTML — walk it, click it, send it
- White-label it under your own brand
PLATE IV — MOCKUPSPECIMEN
IMG 01
WORKING HTML, NOT A WIREFRAME · OPENS IN ANY BROWSER
WHY DEPTH WINS
A blueprint is not a draft, and not a score.
Three ways to start a page. Only one of them starts from what your searchers are actually doing.
One-shot AI writer
Starts from
Your prompt
Structure decided by
The training average
You receive
Text that sounds finished
The reasoning
Hidden in the weights
Who can execute
You — editing, heavily
Optimization checklist
Starts from
Your existing draft
Structure decided by
You, against a score
You receive
A number and a term list
The reasoning
A score, unexplained
Who can execute
Whoever interprets the score
Intent Atlas
Starts from
The live SERP — top 20 captured, fetched, read
Structure decided by
Evidence, argued section by section
You receive
Blueprint · writer's guide · build guide · mockup
The reasoning
Written down, inspectable
Who can execute
Any writer, any builder — the guides carry it
CLIENT WORK
Built for agencies from day one.
The deliverable was designed to be delivered — to a client, under your name, as part of your retainer.
Client workspaces
Every client's projects, runs and packages — cleanly separated, one subscription over all of them.
Seats & roles
Bring the whole team: seats, roles and permissions on every plan, not as an upsell.
White-label exports
Branded packages and PDFs — the research under your name, ready for the client meeting.
No feature gates
Every plan is the whole product. Plans differ in one thing only: credits.
THE AGENCY PAGE — WORKSPACES, WHITE-LABEL, GUEST-POST ECONOMICS →
PRICING
Three plans, one product — the difference is credits.
Credits meter the work: a premium on-page blueprint is 50, standard is 25, guest-posts are 15 and 5. Spend yours on any mix.
Starter
$199 /mo
200 CREDITS / MONTH
4 premium on-page blueprints a month — or 8 standard, or 40 standard guest-posts.
EXTRA CREDITS $1.00 · CREDITS STAY GOOD 12 MONTHS
Start on StarterGrowth
$499 /mo
500 CREDITS / MONTH
10 premium on-page blueprints a month — or 20 standard, or 100 standard guest-posts.
EXTRA CREDITS $0.90 · CREDITS STAY GOOD 12 MONTHS
Start on GrowthScale
$999 /mo
1,000 CREDITS / MONTH
20 premium on-page blueprints a month — or 40 standard, or 200 standard guest-posts.
EXTRA CREDITS $0.80 · CREDITS STAY GOOD 12 MONTHS
Start on ScaleYEARLY: 12 MONTHS FOR THE PRICE OF 9 · YEARLY CREDITS STAY GOOD 2 YEARS
Full pricing & plan comparison →No free tier and no trial theater — the $199 Starter month is the trial: the entire product, 200 credits, cancel anytime.
QUESTIONS
Asked before subscribing.
How is this different from an AI writing tool?
An AI writer gives you finished-sounding text from a standing start. Intent Atlas delivers the decision layer above the writing: what the page must cover, in what order, and why — argued from the pages that actually rank for your keyword. You (or your AI writer) then execute against a plan instead of a guess. The writer's guide makes any writer better; it doesn't replace them.
What exactly does a credit buy?
Runs have fixed prices, shown before you start: a premium on-page blueprint is 50 credits, standard on-page is 25, premium guest-post is 15, standard guest-post is 5. Every run returns the same four-part package. On Starter, one credit is one dollar — the arithmetic is deliberately simple.
How deep is the research, really?
Per run: the top 20 results captured and localized to your market, every ranking page fetched whole, each page broken into sections and read, queries and evidence grouped into an intent map, and a blueprint composed with its reasoning written down. Runs are glass-box — you watch all seven stations work in the app, with live counts of pages fetched, sections read and query groups traced.
Who owns the deliverables?
You do — everything in every package. Publish it, adapt it, rebrand it, deliver it to clients. White-label exports are included on every plan.
What happens if a run fails?
Its credits come back automatically, in full. Delivered packages stay downloadable for the life of your account.
Is there a free trial?
No. The $199 Starter month is the trial: the entire product, 200 credits, no feature gates, cancel anytime. The product's value is the depth of a real run, so the trial is a real month of it.
Do unused credits roll over?
Yes — credits stay usable for 12 months from the month they're granted while your subscription is active (2 years on yearly plans that keep renewing). If you cancel, remaining credits expire at the end of your paid period; everything already delivered stays yours.
Will it work for my niche?
The pipeline makes no assumptions about your vertical. It reads what ranks for your keyword in your market — whatever the keyword, whatever the market — and argues the page from that evidence. If the SERP exists, it can be surveyed.
Put a keyword on the desk.
Pick the plan that matches your volume — most teams start on Growth — or open the sample package first and judge the document.
$199 STARTS IT · CANCEL ANYTIME · FAILED RUNS RETURN THEIR CREDITS
Newly launched — no borrowed logos, no invented case studies. Every account is onboarded personally. About the launch →