How measurement works

A measurement system, not a screenshot collection.

AskAgain controls what is asked, repeats the observation, and preserves the underlying answers so changes can be examined instead of guessed.

How measurement works

01

Configure the subject

Define the company or solution, its market context, known alternatives, proof points, and language.

02

Freeze the question set

Brand projects use eight fixed slots; solution projects use five. Every run stores the exact question-set version.

03

Repeat across platforms

Run the same questions against the selected AI platforms and keep each repetition as separate evidence.

04

Inspect and compare

Review mentions, position, competitors, grounded findings, and the raw answers. Version boundaries make question changes visible.

Two distinct instruments

8

Brand projects

Measure whether and how a company appears across Discovery, Validation, and Decision questions. Eight controlled question slots.

5

Solution projects

Measure one defined offer against its intended problem, audience, requirements, alternatives, and proof. Five controlled slots with separate grounded accuracy analysis.

03 / Cadence

Cadence is part of the evidence

A single answer is an observation, not a trend. Paid levels define both repetitions and the interval between runs. Starting a paid run early consumes the upcoming scheduled occurrence; it never creates an extra run.

04 / Agency

For agencies and client teams

Agency workspaces keep client environments separate while applying one measurement method. Owners approve billable changes; members can manage existing client work without seeing agency payment details.

Frequently asked questions

Does AskAgain claim to know what an AI system thinks?

No. It records observable answers under a controlled setup and keeps the evidence attached.

Can questions change?

Yes, within the fixed instrument structure. A new immutable question-set version is created so reports can show the comparison boundary.

Which platforms are included?

Claude, ChatGPT, Gemini, and Perplexity are selected by default. DeepSeek is available but not selected by default. Platform choice does not change v1 pricing.

Is the sample report a customer case?

No. Falkenwerk Process Cooling is an explicitly fictional composite with internally consistent demonstration data.

What happens after cancellation?

Runs stop immediately. A cancelled project can be deleted now or retained read-only for 60 days. Account cancellation provides 90 days of read-only results and export access after period end.

Can a Free project be replaced repeatedly?

No. Its identity, type, and monitored subject are locked during the first 28 days, and it can run manually only once per 28 days.

Still have a question?

Contact the operating team at support@askagain.io.