Live churn monitor · updated July 2026

Watching an algorithm change
break customer loyalty.

ShopSphere AICV tracks how a single platform algorithm update rippled into a sharp rise in cancellations — and the TabTransformer model built to catch it early.

3.55%
Baseline churn
11.64%
Post-update churn
+8.09pp
AICV change
74.1%
Model ROC-AUC
scroll
01 · Algorithm shock

What the update did to churn

AICV compares the historical churn baseline against churn observed after the ShopSphere algorithm update, isolating the size of the shock.

Historical baseline churn
0%
Pre-update period
Post-update churn
0%
After algorithm change
AICV change
0
Post-update minus baseline
Algorithm update
July 2026
Shock / intervention point
Baseline
3.55%
Post-update
11.64%
0

The post-update churn rate increased substantially compared with the historical baseline, highlighting the business impact associated with the platform algorithm change.

02 · Data

Dataset overview

Longitudinal synthetic customer data created to model ShopSphere customer behavior before and after the algorithm update.

0
Customer-month observations
0
Unique customers
0
Monthly observations per customer
03 · Model

The TabTransformer model

A custom TabTransformer-style neural network was trained to classify customer churn using behavioral, engagement, visibility, pricing and satisfaction features. Metrics below are from a cost-based decision threshold, not the default 0.50 — see why underneath.

Accuracy
0%
Precision
0%
Recall
0%
0%
ROC-AUC

Why accuracy looks low

The decision threshold is chosen to maximize expected retained value, not accuracy — since missing a churner costs an assumed 25× more than an unnecessary retention outreach, the optimal policy flags a much wider net of customers. That trades accuracy for recall on purpose. At a precision of 12.9%, roughly 7 out of every 8 flagged customers are false alarms — a real deployment would need to weigh that against how many customers a retention team can actually reach.

Customer
features
Categorical
embeddings
Transformer
encoder
Numerical
features
Churn
probability

Model configuration

Embedding dimension64
Attention heads4
Transformer layers2
Feed-forward dimension96
Dropout0.15
Decision threshold0.17

Baseline comparison

Logistic Regression and Histogram Gradient Boosting, trained on identical data splits and evaluated with the same cost-based threshold logic.

ModelAccuracyPrecisionRecallF1ROC-AUCThreshold
Logistic Regression 50.9% 14.0% 81.8% 23.9% 73.5% 0.35
Gradient Boosting 14.5% 9.9% 98.8% 17.9% 68.2% 0.02
TabTransformer (deployed) 44.6% 12.9% 84.7% 22.4% 73.3% 0.17

Logistic Regression matches or slightly exceeds the TabTransformer on precision, F1, and ROC-AUC. On a dataset this size (1,000 customers), that's expected — the added architectural complexity isn't earning its keep here. The TabTransformer was retained for this project because it was the assigned architecture, not because the comparison favors it. Gradient Boosting's threshold collapsed to near-zero, flagging almost every customer — a sign its probabilities aren't well-calibrated out of the box, not a genuinely useful policy.

04 · Try it

Estimate churn risk yourself

Loading the trained model weights (~1 MB) for real in-browser inference…

Churn probability — real model output
Waiting for the model to load…

Every field above feeds the actual trained TabTransformer weights via ONNX Runtime, running in your browser. The 16 numerical features not shown here are held at their training-set average for each prediction.

05 · Decision strategy

Business cost & decision strategy

The model supports retention prioritization by balancing the cost of unnecessary interventions against the potential value lost when a churner is missed.

Intervention cost
0
Illustrative assumption
Lost customer value
0
Illustrative assumption
FN cost ratio
0
Potential cost comparison

Drag the decision threshold

0.17
Model default sits at 0.17 · scored by the real trained model on 400 synthetic customers generated in your browser. Since these customers are synthetic, there's no observed outcome for them — each one's "actual churn" is simulated by drawing from that customer's own model-predicted probability, so precision/recall here describe the policy's expected behavior, not a measured result on real data.
Customers flagged
Precision (simulated)
Recall (simulated)
Net expected impact

Business decision

Customers with elevated predicted churn probability can be prioritized for retention interventions. The 0.17 decision threshold maximizes expected net value under the ₹200-vs-₹5,000 cost assumption — but as shown above, at this precision (12.9% on the real held-out test set) most flagged customers will be false alarms. That's a deliberate tradeoff for this cost ratio, not a flaw to hide, and any real deployment should weigh it against retention-team capacity.

Sample customer queue

Synthetic customers, scored by the real model. Flag status updates live as you move the threshold above — click any column to re-sort.

Customer Tenure (mo) Visibility Satisfaction Segment Churn risk Status
06 · Pipeline

End-to-end analytics workflow

ShopSphere
data
AICV
analysis
Power BI
dashboard
TabTransformer
model
Retention
decision
07 · Notes

Project note

This project uses a synthetic dataset created for academic demonstration. The observed relationship between the algorithm update and churn should therefore not be interpreted as causal evidence from a real commercial platform.


Model performance should be evaluated further on real-world customer data before deployment in an operational retention system.