LLM ShelfLLM Shelf

Methodology

From a business question to a decision-ready report.

A process designed to go beyond charts and lead to specific actions.

01

Product and category definition

We define the product role, claims, format, competitors and market context.

02

Demand and user-language analysis

We use search-demand data, category questions and purchase scenarios.

03

Use cases and personas

We build situations in which a user may ask AI for a recommendation.

04

Prompt battery

We create unbranded, branded, comparison and contextual questions.

05

Models and repetitions

We run selected models and measure answer stability.

06

Follow-ups

We examine what happens after the user asks for specific brands and products.

07

KPI scoring

We measure visibility, recommendations, competition, sources and risks.

08

Interpretation

We separate raw data from conclusions and business decisions.

09

Recommendations

We turn the findings into actions for marketing, ecommerce and agency teams.

10

Rerun

After implementation, we measure changes in answers and priorities.