Where does AI search volume come from?

SKYA reads how often a question is asked of AI in the market you track. Where AI demand is not measured for a prompt, it falls back to the search volume of the question's core phrase.

How is Prompt Research different from Prompt Tracking?

Research finds and scores prompts once, before you commit to them. Tracking asks the prompts you keep on a schedule and stores every answer.

How many prompts does one run find?

25 on Starter, 100 on Brand Pro and Agency, and 300 on Agency Pro, each tested before you see it.

Does a research run cost credits?

Yes, one credit per research, from your monthly credits or a pack.

Can I research a topic outside my project?

Yes. Topic mode researches any subject you type, so you can size a new market before you track it.

Prompt Research

Find the questions your buyers ask AI

Prompt Research writes the prompts buyers put to AI about a topic or your project, reads their AI search volume in your market, tests each one in AI answers, and scores which are worth tracking. One click adds the winners.

Start 7-day free trial Which plans include it

Free 7-day trial. No credit card required.

ExampleInvented brand and numbers, drawn from the real screen. Not a customer result.

Northwind Gear/Prompt Research/Discover

ultralight tentsTopicUnited States

Prompts tested

100

this run

Rival named

you are not

Worth tracking

score 70+

best ultralight tent for rain
CommercialTrailForge named
86 Track
TrailForge vs Northwind Gear tents
ComparativeTrailForge named
81 Track
lightest two person backpacking tent
CommercialOpen field
74 Track
how to stop tent condensation
InformationalSummit & Co named
62 Track
where to buy a tent in Denver
TransactionalRidgeline named
55 Track

Decisions it settles

The calls it helps you make, and the move after

Before you spend a tracked prompt, know which questions are worth it.

The questionWhat SKYA showsWhat you do next

01
Which questions should we track?
Every prompt scored on intent, how contested it is, whether rivals appear, its demand and its AI referral signal, so the list sorts itself.
In SKYA: Discover
Add to tracking
02
Where are rivals already being recommended?
The prompts where a rival is named in the tested answer and you are not.
In SKYA: Rivals
Track the prompt
03
Is this question big enough to matter?
AI search volume for the prompt in the market you track, with the search volume of its core phrase as the fallback.
In SKYA: Discover
Sort by demand
04
What should we write next?
Any prompt opens the Content Engine with the question and the evidence already filled in.
In SKYA: Content Engine
Draft the answer

A closer look

A score you can take apart

Every prompt carries one value score, and the parts behind it: buying intent, how contested it is, whether rivals appear, demand in your market and the AI referral signal. A high score is never a black box.

Sort the whole run by value, demand or rival presence
Add a prompt to tracking in one click
Or send it to the Content Engine to write the answer

ExampleInvented brand and numbers, drawn from the real screen. Not a customer result.

Northwind Gear/Prompt Research/Prompt

best ultralight tent for rain

CommercialTrailForge named

Value score

Buying intent
How contested
Rivals present
Demand in your market
AI referral signal

Add to tracking Draft the answer

How it works

Three steps, start to finish

Step 1: Pick a topic or your project
Research a topic you type, or let SKYA start from your project's site, rivals and tracked prompts.
Step 2: SKYA writes and tests the prompts
Up to 300 prompts per run on Agency Pro, each tested in AI answers and scored before you see it.
Step 3: Keep the winners
Add a prompt to tracking in one click, or send it to the Content Engine to write the answer.

What it does

Everything in it, shipping today

Research that ends in a tracked prompt or a draft, not a spreadsheet.

Topic or project mode
Start from any subject, or from what SKYA already knows about your project.
Intent on every prompt
Informational, comparative, commercial, transactional or navigational, so a list of questions reads as a buyer journey.
A value score per prompt
One score from intent, how contested the prompt is, rival presence, demand and AI referral signal.
AI search volume in your market
How often the question is asked of AI where you sell, with search volume as the fallback.
Rivals already named
The tested answer shows which brands AI recommends today, before you track a thing.
One click to tracking or a draft
A good prompt becomes a tracked prompt or a Content Engine brief without retyping it.

Measured, not modelled

Figures you can check

A prompt with no measured AI demand is not given an invented one. It shows the search volume of its core phrase, or no figure.

How every number is calculated

prompts per run on Agency Pro: 30025 on Starter and 100 on Brand Pro and Agency.
countries on Agency Pro: 5Demand is read in the market you track: 1 on Starter, 2 on Brand Pro, and 3 on Agency.
per research run: 1 creditFrom your monthly credits or a pack.