Skip to main content
hitak
Method One answer is not a trend

Measure what AI really says about your brand.

Repeatable, multi-model, documented measurement to understand your presence, your competitors and the sources shaping answers.

Privacy policy

Discover the method

For brand, content and acquisition teams — and their agencies.

AI 1 AI 2 AI 3 Pure AI
Question 042 · “sensitive back” persona Which mattress should I choose for back pain?

For a sensitive back, prioritise firm support and good motion isolation. Your brand is often mentioned for its zoned mattresses, while Competitor 1 offers a more affordable range. Competitor 2 remains an option for side sleepers.

Extracted entities
Your brand #1 Competitor 1 #2 Competitor 2 #3
Cited sources media-1.exemple comparateur-2.exemple +2
Brand mentioned · position 1 Cycle 6 · frozen corpus v3
IGPI · Your brand
41 % ▲ 6 pts
6 comparable cycles
Citation share · cycle 6
  • Your brand 34 %
  • Competitor 1 27 %
  • Competitor 2 18 %
  • Other detected brands 21 %

Fictional data · invented brand

Why hitak

A measurement, not a snapshot.

Every figure retains its evidence, scope and limitations. Included cycles run automatically; no overage is charged without your agreement.

Every finding keeps its evidence.

evidence
Finding · AI 3 · web search Your brand is absent from 7 of 12 answers.
Question 042 · “sensitive back” persona View the 7 answers

Obs. 2 · 8 Sept · 09:14 media-1.exemple … for a sensitive back, Competitor 1 and Competitor 2 are the most frequently recommended options thanks to their firm support. Your brand: absent

Obs. 6 · 8 Sept · 11:40 comparateur-2.exemple Reviews highlight Competitor 1 for its value for money; Competitor 3 comes up for side sleepers. Your brand: absent

Obs. 11 · 8 Sept · 16:05 forum-3.exemple Several users mention Competitor 2 , without mentioning another brand.

Approved once, measured every month.

scope
Included cycle · October
Volume
1,920 answers
Execution
Scheduled batch
Cost
Included in the subscription
Next scheduled cycle · 1 Oct No overage expected.

Data isolated by organisation.

security
Organisation A Brand 1 Brand 2
Organisation B Brand 3 Brand 4

PostgreSQL isolation Separate sessions per surface Context revalidated on every request

A versioned corpus, flagged breaks.

versions
Corpus v1 Corpus v2 Corpus v3
Break flagged

Cycle 6 · corpus v3 Comparable with cycles 4 and 5. Comparison with cycle 3 is stopped, never extrapolated.

Reports ready to present.

reports
Report · cycle 6 Visibility in AI answers Base: 96 observations

PDF Evidence Corpus v3

The product

From your reference framework to actionable measurement.

You configure the brand, its markets, personas and intentions. hitak then freezes the scope before any execution.
  1. 01 Configure the scope Offering, market, competitors, personas and intentions: everything needed for comparable measurement.
  2. 02 Freeze before execution The corpus and variations are locked. Scope is approved before activation.
  3. 03 Observe each regime General-purpose assistants and web-native engines are measured separately, never averaged together.
  4. 04 Connect finding, evidence and action Every report exposes its actual measurement base, evidence and a prioritised action plan.
Your brand / Overview

Visibility in AI answers

Cycle 6 · 8 Sept 2026 Corpus v3 · frozen FR · France

IGPI · pure AI 41 % ▲ 6 pts Average position when mentioned 2,1 across 3 assistants Observed source domains 38 including 4 belonging to the brand

IGPI per cycle · pure AI Cycles 1 to 6

Your brand Competitor 1 Competitor 2

No trend is drawn across the break: different corpus, comparison stopped.

Recommendations no promised uplift

Publish a “sensitive back” guide detailing your support zones ●●● High

Evidence: absent from 7 of 12 answers · web search

Obtain a mention on comparateur-2.exemple ●●○ Medium

Evidence: source cited 14 times, brand never associated

Harmonise range names across the website and listings ●○○ Marginal

Evidence: 3 name variants extracted from answers

Methodology

Depth beats frequency.

One isolated answer is not enough to measure a brand. hitak repeats observations, freezes their conditions and documents their limitations.

Corpus v3 · 8 questions × 12 observations Top 3 Mentioned lower down Absent

back mattress9 / 12

firm or soft7 / 12

best 20266 / 12

side sleeper5 / 12

customer reviews4 / 12

warranty3 / 12

delivery3 / 12

€500 budget2 / 12

IGPI · Your brand 41 % 39 of 96 observations

One answer would say “present” or “absent”. Repeated measurement shows your brand in the top 3 of 9 out of 12 answers for “back mattress”, but only 2 for “€500 budget”.

Repeat without moving the target

The corpus and its variations are frozen so repetitions address the same measurement question.

Separate search regimes

General-purpose assistants and web-native engines are not added together as if they produced the same observation.

State the limitations

An incompatible corpus or scope stops the comparison. A recommendation remains high, medium or marginal: never a predicted uplift.

Solutions

One method, two uses.

For agencies

Scale measurement, not rushed conclusions.

hitak gives consulting teams a shared framework to prepare audits, explain differences and deliver traceable reports.
  • Multiple brands, one stable method
  • Explicit scope before spending
  • Defensible deliverables
Solutions for agencies

For brands

See where your brand stands in AI answers.

hitak helps brand, content and acquisition teams distinguish presence, influence and evidence without treating one answer as a universal truth.
  • Presence and position, per assistant
  • Sources and web consistency
  • A baseline to track changes
Solutions for brands