eCommerce Data Infrastructure: 90% Faster Reporting for Furniture Retailer

Client Background

E Living Furniture is an online furniture retailer with over 5,000 SKUs, generating $12M in annual revenue across their Shopify store and marketplace channels. Despite strong sales, their data was fragmented across multiple systems, making it nearly impossible to get a clear picture of business performance.

The Challenge

E Living faced critical data challenges hampering their growth:

1. Data Silos: Sales data in Shopify, advertising in Google/Meta, inventory in their ERP, and customer data in Klaviyo—none of it connected.

2. Manual Reporting: The operations team spent 15+ hours weekly manually compiling reports from different platforms into spreadsheets.

3. Delayed Decisions: By the time reports were ready, the data was already 3-5 days old, too late for timely pricing or inventory decisions.

4. Pricing Blind Spots: Without real-time margin visibility, they couldn't optimize pricing dynamically based on inventory levels or competitor movements.

5. Attribution Gaps: Marketing couldn't accurately attribute sales to specific campaigns or channels.

Our Solution

We built a comprehensive data infrastructure that unified all business data:

Data Pipeline Architecture

Using Fivetran for reliable data ingestion:

• Shopify orders, customers, and products
• Google Ads and Meta Ads performance data
• Klaviyo email marketing metrics
• ERP inventory levels and costs
• Google Analytics 4 website behavior

Cloud Data Warehouse

BigQuery as the central data repository:

• Automated data modeling with dbt
• Historical data preservation for trend analysis
• Optimized query performance for large datasets
• Cost-effective storage with intelligent tiering

Real-Time Dashboards

Looker Studio dashboards providing:

• Executive KPI overview with daily refresh
• Product performance with margin analysis
• Marketing attribution across all channels
• Inventory health and reorder alerts
• Customer cohort and LTV analysis

Automated Pricing Engine

A custom pricing optimization system:

• Monitors competitor prices via web scraping
• Calculates real-time margins per SKU
• Suggests price adjustments based on inventory velocity
• Alerts team to pricing opportunities

Technical Implementation

Data Flow

Source Systems
├── Shopify (Orders, Products, Customers)
├── Google Ads (Campaigns, Ad Groups, Keywords)
├── Meta Ads (Campaigns, Ad Sets, Ads)
├── Klaviyo (Campaigns, Flows, Subscribers)
├── ERP (Inventory, Costs, Suppliers)
└── GA4 (Sessions, Events, Conversions)
    ↓
Fivetran (Automated Ingestion)
    ↓
BigQuery (Data Warehouse)
    ↓
dbt (Data Transformation)
    ↓
├── Looker Studio (Dashboards)
├── Pricing Engine (n8n + Sheets)
└── Alerts System (Slack Notifications)

Key Technical Decisions

1. Fivetran over Custom ETL: Chose managed connectors for reliability and maintenance-free operation

2. BigQuery for Scale: Serverless architecture handles their growing data volume without infrastructure management

3. dbt for Transformations: Version-controlled SQL models ensure consistent, documented data logic

4. Looker Studio for Accessibility: Free, familiar interface lowered adoption barriers for the team

Results

The data infrastructure transformed E Living's operations:

Reporting Speed

• Before: 3-5 days for comprehensive reports
• After: Real-time dashboards updated hourly
• Improvement: 90% faster access to insights

Operational Efficiency

• 15 hours/week saved on manual reporting
• Zero data entry errors with automated pipelines
• Single source of truth for all business metrics

Business Impact

• 12% margin improvement through dynamic pricing optimization
• 23% reduction in stockouts with inventory alerts
• 18% increase in ROAS through better attribution

Data Quality

• 100% data freshness with automated syncs
• Historical analysis enabled with 2 years of preserved data
• Audit trail for all data transformations

Key Insights

1. Start with Business Questions: We mapped dashboards to specific decisions before building, ensuring relevance.

2. Data Quality First: Invested heavily in validation and cleaning—garbage in, garbage out.

3. Incremental Rollout: Launched core dashboards first, then added complexity based on user feedback.

4. Documentation is Essential: Comprehensive data dictionary ensured team-wide understanding.

Dashboard Highlights

Executive Overview

• Daily revenue, orders, and AOV trends
• Channel mix and margin analysis
• YoY and MoM comparisons
• Key alerts and anomalies

Product Performance

• SKU-level profitability
• Inventory turnover rates
• Price elasticity indicators
• Bestseller and slow-mover identification

Marketing Attribution

• Multi-touch attribution model
• Channel ROAS comparison
• Campaign-level performance
• Customer acquisition cost by source

Long-Term Impact

One year after implementation:

• The data infrastructure processes 2M+ rows daily
• Team makes data-driven decisions in minutes, not days
• Pricing optimization has generated $180K in additional margin
• The system scales automatically with business growth

Conclusion

E Living Furniture's data infrastructure project demonstrates the power of unified analytics for eCommerce. By connecting fragmented data sources into a single, real-time view, we enabled faster decisions, automated manual processes, and unlocked pricing optimization opportunities. The 90% improvement in reporting speed was just the beginning—the real value lies in the business decisions now powered by accurate, timely data.