Kerala Ayurveda prepared for scale with AhaInsights
With an 80-year heritage and millions of loyal customers, Kerala Ayurveda’s 2024 Shopify expansion was built to move fast.
AhaInsights partnered with the data team from the early months, building the high-performance data models and tracking needed to turn massive traffic into actionable growth insights.
Objectives
Single Source of Truth
Create a single source of truth for KPIs like traffic, revenue, sales attribution to paid and organic channels by combining data from Shopify, GA4 and Marketing reports.
Store Tracking for Optimisation
Implement a tracking setup to track website interactions in minor detail, and build data models to simply analysis for recurring tasks like conversion rate, user journeys, etc.
Reliable Automation Infra
Set up data infrastructure that runs fast, adapts to new data, and expands without breaking.
AhaInsights Solution
Granular Tracking
Set up 200+ GTM events to track every micro-interaction. Followed best practices for event names and field types.
Unified data source
Combined GA4 events, Shopify Orders, Meta, and Google Ads in Bigquery from Session -> Sales to create a single source of truth.
Custom Data models
Wrote custom SQL queries in Dataform to combine the data into detailed models ready for analysis and reports.
Stakeholder alignment
Helped the data team get stakeholders alignment by explaining the data gaps and differences.
Results
Aligned stakeholders
With stakeholders aligned on KPIs, team can focus on solving bigger problems than alignment.
Clear Attribution
Confidence in attribution for marketing, affiliate and other sources. Team can scale channels with confidence.
Automated Reporting
Team is able to setup new reports on Metabase and Looker Studio without dealing with messy raw data.
Deep(er) Insights
With a single source of truth, data team can run analysis and build reports not available anywhere else.
Testimonials
Nishant, Head – Data and Analytics
… Aadi structured things so we can quickly answer funnel questions—drop-offs, conversion paths, step-wise performance, and other “what happened and why” investigations—without turning every question into a mini engineering project…