Platform

Online SaaS platform configured to your context

A platform with simple and precise campaign setup. Immediate collection and visualisation of AI answers, for specialised analysis and clear recommendations.

Visual: robot holding a screen that presents the DeepReco platform
Platform

An ultra-configurable platform

A 3-step platform for quickly obtaining precise results: simple setup of your campaigns, intuitive collection and display of AI answers, then specialised analysis and clear recommendations.

  1. 1

    Campaign design

    Build personas, define your targets, and generate real queries at scale.

    Explore step →
  2. 2

    Collect & measure

    Automated multi-engine collection infrastructure across the main AI platforms.

    Explore step →
  3. 3

    Analyse & recommend

    Analyse your positioning and the best-performing sources, then turn the data into a concrete action plan.

    Explore step →
#1 Campaign Design

Configure the solution to your context

Easily generate thousands of prompts tailored to your categories, brands and customer profiles across all AI models.

Start with the objective

Defining an objective gives you a strategic objective, then a concrete result, backed by precise personas and criteria.

Discover our convictions & method
OBJECTIVEStrategicAlignment & scopeLeversTarget AI enginesPersonasCompetitorsCriteriaPrecise resultMeasurable

Automatically generate thousands of prompts representative of your consumers

3,000 ChatGPT prompts

Each customer has their own needs, expectations and decision criteria. Our generator reproduces this diversity by creating personalised prompts that represent the real searches of your customers.

Identify the intents, needs and profiles that drive AI recommendations, and generate thousands of calibrated prompts, ready to deploy.

#2 Collect & Measure

Deploy, capture, and analyze at scale across all major AI platforms

An automated multi-engine collection infrastructure that measures, qualifies, and monitors every AI response in depth.

01

Collect AI responses

Your prompts run on every AI engine. Responses are collected, but also the sources that influence the AI, all timestamped and broken down by country.

Live multi-engine querying● 100% direct
ChatGPT Gemini Perplexity Claude
Prompt · ChatGPTI run 30 km a week on the road, I slightly overpronate and my budget is under €130. Which running shoes would you recommend, and which suit my profile and long distances best?
Réponse collectée

For slight overpronation on road, stability running shoes are generally recommended for long runs: they support the inner arch without stiffening your stride. …

Sources citées
test-running-exemple.fr/comparatif
club-course-exemple.fr/guide.pdf
media-sport-exemple.fr/article
Zéro biais d’historique • Réponses et sources collectées simultanément.
02

Measure recommendation power

For each response, we measure whether the AI recommends your brand through a recommendation score. This score weights your brand’s position, sentiment, presence, number of mentions, the context in which it is cited and the share of voice it obtains.

Result measurement & challenger benchmark● Analyse chirurgicale
Position Score64/100
Royal C.
64
Hill's
58
Purina
47
Sentiment Score72/100
Royal C.
72
Hill's
70
Purina
68
Share of Voice27.8%
  • Royal C.27.8%
  • Hill's21.2%
  • Purina16.8%
Total Mentions384
162
Royal C.
98
Hill's
67
Purina
« Pour une peau sensible, Marque A et Marque B sont souvent citées en premier… »
Extrait de réponse IA, analysé par moteur.
03

Analyze strengths and weaknesses

Understand what AI engines praise, identify or criticize about your brands, products and competitors.

Évaluation experte
AI-expressed sentimentScore global : 58.4 / 100
Votre Marque
Produit A60.8
Critères les plus forts
Digestibilité / Tolérance90
Preuve clinique85
Réseau pro & web82
Critères les plus faibles
Sans céréales / Sans parfum0
Sur-mesure35
Coût relatif42
04

Analyze the types of sources

Analyze the site type, reliability, volume and format.

Source qualification● Indexation validée
Collaborative encyclopaediaen.wikipedia.org/wiki/Pronation
Reliability, expertise & reputationHigh
Scientific consensus & neutralityHigh
Authority & author signatureMultiple contributors
Institutional anchoringCited sources
Volume, depth & exhaustivenessVery high
Volume & topic coverageVery broad
Quantified technical dataHigh
Direct answer to intentDirect
Format, typology & RAG usabilityHigh
Container typeHTML
Tabular structureYes (tables)
Semantic modularityStrong (sections)
#3 Analysis & recommendations

Read the results

Recommendation score

A synthetic indicator out of 100, calculated from several weighted scores: position, sentiment, share of voice and mention volume. It gives a global view of each brand’s performance, makes it possible to compare it with competitors, and pinpoints the factors behind the gaps.

Score #1
48,6
Royal Canin1
42,0
Hill's2
33,9
Purina Pro3
20,8
Virbac4
7,5
Edgard & C.5
Position Score64/100
Royal C.
64
Hill's
58
Purina
47
Virbac
31
E&C
22
Sentiment Score72/100
Royal C.
72
Hill's
70
Purina
68
Virbac
61
E&C
55
Share of Voice27.8%
  • Royal C.27.8%
  • Hill's21.2%
  • Purina16.8%
  • Others34.2%
Total Mentions384
162
Royal C.
98
Hill's
67
Purina
35
Virbac
22
E&C

Analysis by personas

Refine the analysis by skin type, age, severity and affected area. Each profile changes the recommendation score and the brand ranking.

Recommendation score
78/100
Brand A1
62/100
Your brand2
55/100
Brand B3
41/100
Brand C4
36/100
Brand D5

Share of voice

Each brand’s share of citations in AI responses, along with the top 3 market concentration.

#1
LEADERRoyal Canin 27.8%
⬡
TOP 365.8% marché
27.8%Leader
1Royal Canin
27.8%
2Hill's
21.2%
3Purina Pro
16.8%
●Autres
34.2%

By source type

Source classification by type, to see which category feeds AI answers the most.

GEO Score 89
  • Brand.com35%
  • E-commerce30%
  • Forums & authority17%
  • Digital media8%
  • Medical journals7%
  • Others3%

Top sites

Which sites are used the most and influence the AI the most in its answers.

www.reddit.com
1833
www.royalcanin.com
447
vetcenter.purina.fr
352
www.lacompagniedesanimaux.com
351
fr.wikipedia.org
343

Source quality

Source qualification to structure sources and give each one a quality score.

12,902 sources rated out of 14,710 · 87.7% coverage
Reliability, Expertise & Reputation4.8 / 9
Volume, Depth & Exhaustiveness5.7 / 9
Format, Typology & RAG usability5.4 / 9

Source list

Four source examples, two shown at a time, with their type, URL, appearance rate, mentions and qualification scores.

Collaborative encyclopaedia
42 %appearance rate
318mentions
en.wikipedia.org/wiki/Pronation
Reliability, expertise & reputationHigh
Scientific consensus & neutralityHigh
Authority & author signatureMultiple contributors
Institutional anchoringCited sources
Volume, depth & exhaustivenessVery high
Volume & topic coverageVery broad
Quantified technical dataHigh
Direct answer to intentDirect
Format, typology & RAG usabilityHigh
Container typeHTML
Tabular structureYes (tables)
Semantic modularityStrong (sections)
PDF document · tabular content
17 %appearance rate
96mentions
institut-exemple.fr/comparatif.pdf
Reliability, expertise & reputationHigh
Scientific consensus & neutralityHigh
Authority & author signatureNamed authors
Institutional anchoringInstitution
Volume, depth & exhaustivenessHigh
Volume & topic coverageBroad
Quantified technical dataVery rich
Direct answer to intentLate
Format, typology & RAG usabilityModerate
Container typePDF
Tabular structurePartial
Semantic modularityMedium

Expert sentiment

The criteria that drive your strengths and weaknesses: what the AI praises and what holds you back.

Gastrointestinal Chien60.8
TOP CRITÈRES
Digestibilité90
Thérapeutique85
FLOP CRITÈRES
Sans céréales0
Sur-mesure35
Labrador Adulte58.1
TOP CRITÈRES
Réseau pro & web88
Attractivité78
FLOP CRITÈRES
Sans céréales0
Thérapeutique10

Direct head-to-head

A side-by-side duel against your top rival, category by category, to see where you win and where you lose.

Royal Canin28% (49 pts)
VS
Hill's21% (42 pts)
29%
Chats
23%
24%
Chiens
32%
28%
Alimentation
22%

Heatmap map

Each brand’s share of voice broken down by audience segment, displayed as a visual heatmap.

Segments ciblesRoyal CaninHill'sPurina Pro PlanVirbacEdgard & Cooper
Bengal25%25%17%5%9%
Maine Coon35%22%21%5%6%
Ragdoll38%20%26%6%3%
Siamois34%18%16%27%0%

Evolution over time

Track how your performance moves over time and measure the impact of the actions you have put in place.

Action launched25.6 %19.9 %40 %
JanFebMarAprMayJun
Brand ABrand BYour brand
Customisation

A 100% customisable solution

A fast 3-step platform, configured to your needs: your measures, source characteristics, personas, brands, competitors and products.

Your criteria, your rules

Each campaign is set up to match your organisation and objectives, with no fixed template.

● KPIs & VisibilityCritère 01 / 07

Measures

Precisely define the performance indicators tracked across every response generated by AI engines.

Presence01
Brand mentions
Exemple

Your brand cited and indexed in the AI answer

Ranking02
Recommendation position
Exemple

Recommended #1, in top 3, or unmentioned

Perception03
Tone & Sentiment
Exemple

Favourable, neutral, or cautions raised

See the platform on your own brands.

Request a campaign