Skip to content

Questions and Answers

See, question by question, where you lose

The question scoreboard puts every registered question's result on every engine side by side. Open a cell and read the answer itself and the sources it leaned on.

Check technical readinessGet a quote

Free technical readiness check · no sign-up · no card

Question scoreboard

Example
Question scoreboard
QuestionChatGPTGeminiPerplexityClaudeAI Overview
Which software would you recommend for corporate teams?✓✓◆✓◐
Which brands offer payment in instalments?◆◆◆△◆
Is Sample Brand A reliable?✓◐✓◆✓
What is the best option for a small business?◆✓△◆◆
  • named
  • measured, not named
  • could not measure
Example view · drawn with a sample brand and sample data, not a real measurement.

Answers are collected on 5 engines, each one reported on its own:

  • ChatGPT
  • Gemini
  • Perplexity
  • Claude
  • Google AI Overview

Question scoreboard

Example
Question scoreboard
QuestionChatGPTGeminiPerplexityClaudeAI Overview
Which software would you recommend for corporate teams?✓✓◆✓◐
Which brands offer payment in instalments?◆◆◆△◆
Is Sample Brand A reliable?✓◐✓◆✓
What is the best option for a small business?◆✓△◆◆
  • named
  • measured, not named
  • could not measure
Example view · drawn with a sample brand and sample data, not a real measurement.

01

Every registered question, every engine, one table

See at a glance the questions that named you, those that named a competitor, and those that named no brand at all. Cells that could not be measured are marked apart.

  • The table reads from the questions you lose: who you lost to and what the engine read instead sits beside each.
  • Every cell opens to its answers; the summary never loses its evidence.

The answer itself

Example
Perplexitynamed

“Options that stand out for corporate teams include Sample Brand A and Sample Brand B. Sample Brand A is mentioned especially for multi-branch support…”

Sources this answer cited

  • samplebrand-a.comyour site
  • sample-guide.comguides and comparisons
  • sample-forum.netforum
Example view · drawn with a sample brand and sample data, not a real measurement.

02

Filter, read and share every answer

Answers filter by engine, question type, audience, mention state and text search. The filter you choose is written into the address, so you send the same view to your team with a link.

  • Each answer states its state plainly: named, measured but not named, not measured, or a namesake company.
  • Where sources were collected, they are listed under the answer; where they were not, that is said too.

Named alongside you

Example
  • Sample Brand Bin price questionsnamed together
  • Sample Brand Cin trust questionsnamed instead of you
  • Sample Brand Din comparison questionsnamed together

Being named together is an observation; it does not say why the engine chose.

Example view · drawn with a sample brand and sample data, not a real measurement.

03

Who was named in your questions instead of you

For each registered question you see which brands appeared in the answers, what the answer said about them and which sources it leaned on. Quotes are the answer's own words.

  • Questions where a competitor was named and you were not are counted apart.
  • Questions left ambiguous, and those naming no brand, are kept separate.

What the engine searched while answering

Example
  • corporate accounting software comparison
  • e-invoice compatible software recommendation
  • accounting software instalments

Only engines that report their searches share them.

Example view · drawn with a sample brand and sample data, not a real measurement.

04

What AI searched the web for while answering

On engines that search the web and report it, the phrases searched while answering are recorded. The most frequent searches are a list of questions content can answer.

  • Only engines that report their searches share them; the report names which.
  • Whether your brand's name was searched is shown separately.

Period comparison

Example

Previous measurement

Same question set, same version

This measurement

Same question set, same version

Two measurements sit side by side only if they used the same questions; a difference within noise reads 'no change detected'.

Example view · drawn with a sample brand and sample data, not a real measurement.

05

The same questions, comparable measurements

The question set stays fixed from one measurement to the next. Adding a question proposes a new version of the set; two measurements sit side by side only if they used the same questions.

  • Promoting a question from the pool does not change the current measurement; it proposes a new version.
  • Two measurements made with different question sets are not compared.

How it works

From the table to the answer, from the answer to the source

  1. Read the table

    Find the questions you lose and who you lose them to.

  2. Open the answer

    Read the engine's words, the brand it named instead of you and the sources it leaned on.

  3. Set the order

    Lost questions tie into opportunity and action cards, where your team runs the order of work.

Method promise

Fixed questions, results tied to the answers

The question scoreboard shows only the answers to the registered question set. It does not measure how often a question is asked in real life, and it does not say why an engine chose a brand.

Read the method

What we measure

  • Every question's result on every engine
  • The brands named instead of you
  • The full answer text and its sources
  • The searches an engine ran while answering, where it reports them

What we do not measure

  • How often a question is asked of AI assistants
  • The reason behind an engine's choice
  • Change across different question sets

Other modules (9)

One measurement, answers to other questions

Frequently asked questions

Who decides the question set?

The question set is recorded in writing, with a version, before the measurement; we scope it together. It does not change once the measurement starts; a change is proposed as a new version.

Can I add my own question?

You can promote a question from the question pool, which proposes a new version of the set. The current measurement and report do not change; the new question is measured in the next version.

Why do results differ across engines?

Engines read different sources and answer the same question differently. So each engine's result is shown apart, and engines are not ranked against each other when their answer counts differ.

How do I share a view with my team?

The filters you choose in the answer explorer are written into the address. A teammate you send the link to opens the same view, provided they have access to the brand.

See the questions you lose

The free check measures your site's technical readiness. For a measurement with a registered question set, let us scope it together.

Check technical readinessGet a quote

Free technical readiness check · no sign-up · no card