Data, Transformation, and Stack
Nabler and IntuitioLab LCC worked to secure three primary data sources were collected to
perform the analysis including subscribers, buyers, and mobile app users. The subscription
data for last 36 months was extracted out of Recharge. Buyers for last 12 months were
extracted out of Shopify. And the app data was extract as a CSV file.
All three data sets were hosted in AWS S3. Python scripts on AWS EC2 were used to deduplicate
and cleanse data to generate the requested file for third party data augmentation. This
compute environment was also used to analyze customers, their transactions, and build RFM
The data exchange with third-party data vendor was completed via CSV file, and the
augmented file was hosted back in S3. The EC2 compute engine was used for unsupervised
clustering on third party data.
The final augmented file, the RFM segments, and the unsupervised clustering output, was
uploaded into PowerBI, Klaviyo for emailing, and in Glew.io, for online retargeting and
Analytics Outcomes Delivered
As part of third party data augmentation,
Short Par 4 was provided with 380+
additional attributes to enhance the
understanding of existing subscribers and
buyers. The augmentation allowed the
company to better understand a
customer’s typical demographic profile, life
stage, and personal interests.
This information proved useful for Short Par
4, allowing them to fine-tune marketing
content and messaging strategy. Further,
the company learned the key economic
indicators and shopping behavior of their
customers, which will help align products,
sales, and pricing strategy.
The cluster profiles defined their uniqueness and potential marketing action, which can be
personalized for each cluster. Each cluster profile explained the typical demographic, interest
areas, economic purchase power, and shopping behavior.
These clusters were created using Machine Learning with transaction and third-party data.
They included profiles to explain the typical demographic, interest areas, economic
purchase power, and shopping behavior. Eventually, they were grouped into the three core
audience segments that enabled teams including Creative/Content and the Agency
(IntuitioLabs) to use in various marketing programs for audience creation and targeting
Business Insights Delivered
A Microsoft PowerBI dashboard was created to understand key business metrics to
further aid decision-making, and contextualize recommendations.
The table above explains how customer segments were defined. The chart below breaks
downs the customers into VIPs, or how much they exhibit high loyalty and profitability.
This data was analyzed and sliced to produce additional views and generate the insights
such into when VIPs are most likely to shop and purchase.
The following graphs, which are part of the Power BI dashboard, helped the client understand
these VIP customers even more by separating them into different metrics including
relationship length, order size, seasonality, and economic profile.
Marketing Transformation At Short PAR 4