113 Industries - CRM
113 Industries - CRM
Consumer conversations hold valuable insights into the real-world experiences with Continuous Glucose Monitoring (CGM) devices. This project leverages NLP and sentiment analysis to transform unstructured online feedback into actionable insights for product and strategy teams.
VIsion & Problem Statement
VIsion & Problem Statement
Vision:
To empower 113 Industries with a structured analytics framework that surfaces meaningful consumer insights from online conversations, supporting data-driven client recommendations in healthcare and medical devices.
Problem Statement:
113 Industries aims to provide CGM manufacturers with deep, non-obvious consumer insights beyond traditional surveys. However, raw consumer conversations across social media and forums are noisy, fragmented, and unstructured.
Vision:
To empower 113 Industries with a structured analytics framework that surfaces meaningful consumer insights from online conversations, supporting data-driven client recommendations in healthcare and medical devices.
Problem Statement:
113 Industries aims to provide CGM manufacturers with deep, non-obvious consumer insights beyond traditional surveys. However, raw consumer conversations across social media and forums are noisy, fragmented, and unstructured.
VIsion & Problem Statement
Vision:
To empower 113 Industries with a structured analytics framework that surfaces meaningful consumer insights from online conversations, supporting data-driven client recommendations in healthcare and medical devices.
Problem Statement:
113 Industries aims to provide CGM manufacturers with deep, non-obvious consumer insights beyond traditional surveys. However, raw consumer conversations across social media and forums are noisy, fragmented, and unstructured.
Product Goal
Product Goal
Develop a scalable consumer analytics solution to enable 113 Industries to efficiently identify key consumer sentiments, pain points, and competitive differentiators across major CGM brands by extracting structured sentiment, topic themes, and competitive benchmarks to inform client product innovation and positioning strategies.
Develop a scalable consumer analytics solution to enable 113 Industries to efficiently identify key consumer sentiments, pain points, and competitive differentiators across major CGM brands by extracting structured sentiment, topic themes, and competitive benchmarks to inform client product innovation and positioning strategies.
Product Goal
Develop a scalable consumer analytics solution to enable 113 Industries to efficiently identify key consumer sentiments, pain points, and competitive differentiators across major CGM brands by extracting structured sentiment, topic themes, and competitive benchmarks to inform client product innovation and positioning strategies.
User Stories
User Stories
Title | As a/an | I want to | So that |
|---|---|---|---|
Extract Consumer Sentiment | 113 Industries Research Analyst | Analyze consumer perceptions towards CGM brands | Deliver data-backed strategic recommendations to clients |
Benchmark Brand Performance | 113 Industries Strategy Lead | Compare consumer sentiment across major CGM competitors | Identify brand-specific strengths and opportunities |
Identify Emerging Topics | Innovation Consultant | Discover new trends and concerns in consumer conversations | Advise clients on innovation and R&D prioritization |
Attribute-Based Analysis | Insights Analyst | Track sentiment across quality, cost, and ease of use attributes | Provide targeted improvement areas for client products |
Visualize Strategic Insights | Client Delivery Manager | Present insights via clean, executive-ready dashboards | Help clients make faster, more confident decisions |
Title | As a/an | I want to | So that |
|---|---|---|---|
Extract Consumer Sentiment | 113 Industries Research Analyst | Analyze consumer perceptions towards CGM brands | Deliver data-backed strategic recommendations to clients |
Benchmark Brand Performance | 113 Industries Strategy Lead | Compare consumer sentiment across major CGM competitors | Identify brand-specific strengths and opportunities |
Identify Emerging Topics | Innovation Consultant | Discover new trends and concerns in consumer conversations | Advise clients on innovation and R&D prioritization |
Attribute-Based Analysis | Insights Analyst | Track sentiment across quality, cost, and ease of use attributes | Provide targeted improvement areas for client products |
Visualize Strategic Insights | Client Delivery Manager | Present insights via clean, executive-ready dashboards | Help clients make faster, more confident decisions |
User Stories
Title | As a/an | I want to | So that |
|---|---|---|---|
Extract Consumer Sentiment | 113 Industries Research Analyst | Analyze consumer perceptions towards CGM brands | Deliver data-backed strategic recommendations to clients |
Benchmark Brand Performance | 113 Industries Strategy Lead | Compare consumer sentiment across major CGM competitors | Identify brand-specific strengths and opportunities |
Identify Emerging Topics | Innovation Consultant | Discover new trends and concerns in consumer conversations | Advise clients on innovation and R&D prioritization |
Attribute-Based Analysis | Insights Analyst | Track sentiment across quality, cost, and ease of use attributes | Provide targeted improvement areas for client products |
Visualize Strategic Insights | Client Delivery Manager | Present insights via clean, executive-ready dashboards | Help clients make faster, more confident decisions |
Core Features
Core Features
Feature | Description | Priority |
|---|---|---|
Sentiment Analysis Dashboard | Visualizes consumer sentiment across CGM brands and product attributes | P1 |
Topic Modeling (LDA) | Identifies key discussion topics and emerging consumer concerns | P1 |
Attribute-Based Sentiment Scoring | Maps sentiment to attributes like price, usability, quality, and support | P1 |
Competitive Benchmarking | Benchmark sentiment scores and engagement across Dexcom, Libre, Medtronic, Sand enseonics | P1 |
Word Cloud Generation | Highlights common terms and trends discussed by consumers | P2 |
Data Cleaning & Preprocessing | Structuring noisy, unstructured text data for analysis | P2 |
Consumer Demographic Analysis | Segment findings by user traits where available (gender, location) | P2 |
Feature | Description | Priority |
|---|---|---|
Sentiment Analysis Dashboard | Visualizes consumer sentiment across CGM brands and product attributes | P1 |
Topic Modeling (LDA) | Identifies key discussion topics and emerging consumer concerns | P1 |
Attribute-Based Sentiment Scoring | Maps sentiment to attributes like price, usability, quality, and support | P1 |
Competitive Benchmarking | Benchmark sentiment scores and engagement across Dexcom, Libre, Medtronic, Sand enseonics | P1 |
Word Cloud Generation | Highlights common terms and trends discussed by consumers | P2 |
Data Cleaning & Preprocessing | Structuring noisy, unstructured text data for analysis | P2 |
Consumer Demographic Analysis | Segment findings by user traits where available (gender, location) | P2 |
Core Features
Feature | Description | Priority |
|---|---|---|
Sentiment Analysis Dashboard | Visualizes consumer sentiment across CGM brands and product attributes | P1 |
Topic Modeling (LDA) | Identifies key discussion topics and emerging consumer concerns | P1 |
Attribute-Based Sentiment Scoring | Maps sentiment to attributes like price, usability, quality, and support | P1 |
Competitive Benchmarking | Benchmark sentiment scores and engagement across Dexcom, Libre, Medtronic, Sand enseonics | P1 |
Word Cloud Generation | Highlights common terms and trends discussed by consumers | P2 |
Data Cleaning & Preprocessing | Structuring noisy, unstructured text data for analysis | P2 |
Consumer Demographic Analysis | Segment findings by user traits where available (gender, location) | P2 |
Success Metrics
Success Metrics
Metric | Description |
|---|---|
Attribute Sentiment Scores | Sentiment strength and polarity are linked to product attributes |
Topic Coherence Score | Coherence of automatically generated discussion topics |
Brand-Specific Engagement Volume | Number of mentions/posts per brand |
Sentiment Polarity Distribution | Percentage breakdown of positive, neutral, and negative mentions |
Unique Product Pain Points Identified | The number of strategic issues surfaced for innovation input |
Metric | Description |
|---|---|
Attribute Sentiment Scores | Sentiment strength and polarity are linked to product attributes |
Topic Coherence Score | Coherence of automatically generated discussion topics |
Brand-Specific Engagement Volume | Number of mentions/posts per brand |
Sentiment Polarity Distribution | Percentage breakdown of positive, neutral, and negative mentions |
Unique Product Pain Points Identified | The number of strategic issues surfaced for innovation input |
Success Metrics
Metric | Description |
|---|---|
Attribute Sentiment Scores | Sentiment strength and polarity are linked to product attributes |
Topic Coherence Score | Coherence of automatically generated discussion topics |
Brand-Specific Engagement Volume | Number of mentions/posts per brand |
Sentiment Polarity Distribution | Percentage breakdown of positive, neutral, and negative mentions |
Unique Product Pain Points Identified | The number of strategic issues surfaced for innovation input |
Technical Stack
Technical Stack
Models: VADER Sentiment Analysis, Latent Dirichlet Allocation (LDA) for Topic Modeling
Frameworks: Python (NLTK, pandas, scikit-learn, gensim), Tableau, Seaborn, Matplotlib
Data Sources: Public social media platforms, online forums, blogs
Outputs: Attribute-based sentiment scores, topic models, brand benchmarking dashboards, executive insight reports
Models: VADER Sentiment Analysis, Latent Dirichlet Allocation (LDA) for Topic Modeling
Frameworks: Python (NLTK, pandas, scikit-learn, gensim), Tableau, Seaborn, Matplotlib
Data Sources: Public social media platforms, online forums, blogs
Outputs: Attribute-based sentiment scores, topic models, brand benchmarking dashboards, executive insight reports
Technical Stack
Models: VADER Sentiment Analysis, Latent Dirichlet Allocation (LDA) for Topic Modeling
Frameworks: Python (NLTK, pandas, scikit-learn, gensim), Tableau, Seaborn, Matplotlib
Data Sources: Public social media platforms, online forums, blogs
Outputs: Attribute-based sentiment scores, topic models, brand benchmarking dashboards, executive insight reports
Key Results
Processed over 37,000 consumer-generated posts across CGM products
Identified major pain points and brand differentiators by attribute and sentiment
Delivered brand-specific consumer insights to support 113 Industries' client advisory projects
Validated the feasibility of automating consumer insight generation for healthcare product categories
Constraints, Risks, and Mitigations
Constraint / Risk | Impact | Mitigation Strategy |
|---|---|---|
Small volume for emerging brands | Less robust insights for smaller competitors | Adjust analytical weighting and provide context |
Data Noise and Irrelevant Mentions | Reduces the precision of sentiment analysis | Advanced preprocessing and validation sampling |
Sarcasm Detection in Sentiment Models | Potential misinterpretation of polarity | Future model refinement (ensemble or custom lexicons) |
API/Platform Data Limitations | Restricted access to new consumer data | Expand to multi-platform data sources when feasible |
Business Impact
Strengthens 113 Industries’ ability to provide deep, non-obvious consumer insights to healthcare clients
Enables faster turnaround on strategic innovation briefs and competitive analyses
Reduces manual research costs by automating large-scale consumer conversation analysis
Enhances client perception of 113 Industries as a data-driven, innovation-first consulting partner
Provides a scalable framework for expanding similar projects across adjacent healthcare markets
Future Roadmap
Short-Term
Expand to include multi-language consumer feedback (Spanish, French)
Increase the granularity of attribute-based sentiment scoring
Mid-Term
Predict consumer sentiment trends to anticipate product reputation risks
Integrate additional social platforms for richer consumer insights
Long-Term
Build proprietary consumer sentiment libraries by device category
Offer continuous monitoring dashboards to clients as a subscription service
Key Results
Processed over 37,000 consumer-generated posts across CGM products
Identified major pain points and brand differentiators by attribute and sentiment
Delivered brand-specific consumer insights to support 113 Industries' client advisory projects
Validated the feasibility of automating consumer insight generation for healthcare product categories
Constraints, Risks, and Mitigations
Constraint / Risk | Impact | Mitigation Strategy |
|---|---|---|
Small volume for emerging brands | Less robust insights for smaller competitors | Adjust analytical weighting and provide context |
Data Noise and Irrelevant Mentions | Reduces the precision of sentiment analysis | Advanced preprocessing and validation sampling |
Sarcasm Detection in Sentiment Models | Potential misinterpretation of polarity | Future model refinement (ensemble or custom lexicons) |
API/Platform Data Limitations | Restricted access to new consumer data | Expand to multi-platform data sources when feasible |
Business Impact
Strengthens 113 Industries’ ability to provide deep, non-obvious consumer insights to healthcare clients
Enables faster turnaround on strategic innovation briefs and competitive analyses
Reduces manual research costs by automating large-scale consumer conversation analysis
Enhances client perception of 113 Industries as a data-driven, innovation-first consulting partner
Provides a scalable framework for expanding similar projects across adjacent healthcare markets
Future Roadmap
Short-Term
Expand to include multi-language consumer feedback (Spanish, French)
Increase the granularity of attribute-based sentiment scoring
Mid-Term
Predict consumer sentiment trends to anticipate product reputation risks
Integrate additional social platforms for richer consumer insights
Long-Term
Build proprietary consumer sentiment libraries by device category
Offer continuous monitoring dashboards to clients as a subscription service
Key Results
Processed over 37,000 consumer-generated posts across CGM products
Identified major pain points and brand differentiators by attribute and sentiment
Delivered brand-specific consumer insights to support 113 Industries' client advisory projects
Validated the feasibility of automating consumer insight generation for healthcare product categories
Constraints, Risks, and Mitigations
Constraint / Risk | Impact | Mitigation Strategy |
|---|---|---|
Small volume for emerging brands | Less robust insights for smaller competitors | Adjust analytical weighting and provide context |
Data Noise and Irrelevant Mentions | Reduces the precision of sentiment analysis | Advanced preprocessing and validation sampling |
Sarcasm Detection in Sentiment Models | Potential misinterpretation of polarity | Future model refinement (ensemble or custom lexicons) |
API/Platform Data Limitations | Restricted access to new consumer data | Expand to multi-platform data sources when feasible |
Business Impact
Strengthens 113 Industries’ ability to provide deep, non-obvious consumer insights to healthcare clients
Enables faster turnaround on strategic innovation briefs and competitive analyses
Reduces manual research costs by automating large-scale consumer conversation analysis
Enhances client perception of 113 Industries as a data-driven, innovation-first consulting partner
Provides a scalable framework for expanding similar projects across adjacent healthcare markets
Future Roadmap
Short-Term
Expand to include multi-language consumer feedback (Spanish, French)
Increase the granularity of attribute-based sentiment scoring
Mid-Term
Predict consumer sentiment trends to anticipate product reputation risks
Integrate additional social platforms for richer consumer insights
Long-Term
Build proprietary consumer sentiment libraries by device category
Offer continuous monitoring dashboards to clients as a subscription service
More Works More Works
More Works More Works
GOOGLE - INTERGRAM
GEN AI + ML + HACKATHON
2024
2024
GOOGLE - INTERGRAM
GEN AI + ML + HACKATHON
2024
2024
GOOGLE - INTERGRAM
GEN AI + ML + HACKATHON
2024
2024
GOOGLE - INTERGRAM
GEN AI + ML + HACKATHON
2024
2024
TCS - MONOVIZ
ADAS + ML + COMPUTER VISION
2024
2024
TCS - MONOVIZ
ADAS + ML + COMPUTER VISION
2024
2024
TCS - MONOVIZ
ADAS + ML + COMPUTER VISION
2024
2024
TCS - MONOVIZ
ADAS + ML + COMPUTER VISION
2024
2024