![case study](https://cdn.quantiphi.com/2022/11/user-segmentatio-b.png)
User Segmentation and Profiling on Global Microblogging Platform
Media & EntertainmentBusiness Impacts
Data-driven identification of high-value user accounts
Notable increase in user engagement
Individual focus on different user groups
Customer Key Facts
- Country : United States
- Size : 7500+ Employees
- Industry : Social Media
- About : The client is a leading global
social media platform that allows users
to make posts using limited number of words
Determine high-performing social media users and posts
The California-based global microblogging and social networking website company wanted to segment its users based on certain markers and attributes that drive the growth of an account or a post.
Challenges
- Inconsistency in variables
- Difficulty in identifying accurate drivers and attributes to predict superusers
- Identifying a model that guarantees the accuracy of data
![](https://cdn.quantiphi.com/2024/03/88addac4-challenge-1.png)
Technologies Used
![Google BigQuery](https://cdn.quantiphi.com/2024/02/BigQuery-e1656078051285.png)
Google BigQuery
![Google Cloud Function](https://cdn.quantiphi.com/2024/02/Cloud-Functions-e1656078444597.png)
Google Cloud Function
![Google Cloud Storage](https://cdn.quantiphi.com/2024/02/Cloud-Storage.png)
Google Cloud Storage
![Google Cloud Dataflow](https://cdn.quantiphi.com/2024/02/Cloud-Dataflow.png)
Google Cloud Dataflow
![Kubeflow](https://cdn.quantiphi.com/2024/02/Kubernetes-Engine.png)
Kubeflow
Solution
Quantiphi developed an end-to-end solution leveraging the clustering model to segment users with similar behavioral attributes into groups.
Extensive driver's analysis was performed to identify the metrics and attributes that have a propensity to drive post and follower growth.
A custom UI was then built to display different clusters that were created using the clustering model.
Results
- Identifying high-value user accounts for better targetting of customers
- Grouping users based on custom requirements
- Significant increase in user engagement