Don’t writethat book.
Not until the evidence says readers need it. AuthorNX tests nonfiction ideas against real reader signals, argues against them, and tells you plainly which books are worth writing.
Start with an idea, or none.
Explore my idea
Bring a subject, a hunch or a working title. AuthorNX reads it several ways and tests each reading.
Your idea
Money habits for freelancers
Readings to test
- First-year freelancers with uneven income
- Creatives who dread tax season
- Freelancers paying off debt on irregular pay
Sample shown for illustration
Discover for me
Set a market, a budget and a cadence. The Discovery Radar finds reader signals and investigates them.
Discovery Radar
- Market
- Practical small business, US
- Cadence
- Weekly
- Paid research
- Off, $0 a month
- Status
- Paused until you switch it on
Both paths run through exactly the same evidence, scoring, cost accounting and approval steps.
Read the receipt.
One pipeline, five stages. Every call to a source is logged with what it cost and what it found, so you can see exactly how AuthorNX reached its verdict.
Then the decision is yours: approve, reject, watch, or ask for more research. Nothing is produced, spent or published without you.
Research receipt
Sample mission, free sources only. Not real research.
Reconnaissance
Wikipedia topic attentionFree
Stack Exchange reader questionsFree
Open Library catalog checkFree
Objectives
Each with a way to disprove itFree
Diligence
Crossref source checkFree
Google and Amazon demandSkipped
The Skeptic’s case againstFree
Evidence gate
Measured demandUnknown
Promising, unvalidated$0.00
Ten capabilities, shown as they are.
Status comes from tests, not intentions. Pick one to see what works today and what is still missing.
Human- or AI-led missions: reconnaissance, objectives, independent diligence, ranked go / hold / reject.
Partly working. Both modes run end to end on free sources and fixtures. Live Google and Amazon data waits on DataForSEO funding.
Open in Mission Control
Unknown stays unknown.
Many research tools fill gaps with confident numbers. When AuthorNX has not measured something it says so, and an idea can’t score above “promising, unvalidated” until real demand data arrives. Left, a guessed search volume. Right, what AuthorNX shows instead.
Evidence, sample opportunity
- Reader questions
- Observed on Stack Exchange
- Search demand
- Unknown, not measured
- Amazon competition
- Unknown, not measured
- Verdict
- Promising, unvalidated
12k
?
Every source, named.
Free Research
| Source | What it tells AuthorNX | Cost |
|---|---|---|
| Wikipedia | How much attention a topic gets | Free |
| Stack Exchange | Real questions readers ask | Free |
| Open Library | Which books already exist | Free |
| Crossref | Sources and DOIs behind claims | Free |
| Your KDP reports | Your own sales, imported by you | Free |
Sandbox
| Source | What it tells AuthorNX | Cost |
|---|---|---|
| DataForSEO Sandbox | Dummy data in the real shape, never evidence | Free |
Real Data
| Source | What it tells AuthorNX | Cost |
|---|---|---|
| DataForSEO Google | Search demand for keywords | Capped |
| DataForSEO Amazon | Amazon demand and listings | Capped |
| OpenRouter models | Drafting, labeled as model inference | Capped |
Spend nothing by default.
$0
Paid research is off until you switch it on. Every billable call then reserves its maximum cost against mission, daily and account caps, and is refused when the cost is unknown.
The pilot’s default budget setting.
Plain answers.
Does AuthorNX invent numbers?
No. It never shows invented search volumes, sales, rankings or confidence. Anything it has not measured is marked unknown, and AI output is labeled as model inference, never as evidence.
Will it spend money without asking?
No. Paid research is off by default. When you switch it on, every billable call reserves its maximum cost against mission, daily and account caps first, and is refused when the cost is unknown.
Does it publish books for me?
No. AuthorNX never uploads to KDP, changes a listing, buys advertising or publishes anything. People do those steps after they approve them.
What does it use when paid data is off?
Free public sources: Wikipedia, Stack Exchange, Open Library and Crossref, plus the KDP reports you import. Without measured demand the best verdict it gives is “promising, unvalidated”.
What is Sandbox data for?
Testing. DataForSEO’s free Sandbox returns dummy data with the real response shape. It is labeled everywhere it appears and never counts as evidence.
Who is it for?
Right now it is an internal pilot for Jollof Pages, not a public product. It is built to find practical nonfiction books worth writing, and to say so plainly when one is not.