Order cancellations on the Personal Care Online Retailer platform have increased by 20%. The platform allows cancellations within a 24-hour window, but this spike is specifically seen after the 24-hour mark—when products are prepared for shipping.

Key Observations & Clarifications:

• The spike in cancellations is sudden and occurred within the past week.

• It is specific to the app (not the website).

• Primarily affects makeup products, not skincare.

• Payment method (e.g., COD or EMI) is not a factor.

• No recent external changes (e.g., competition, pricing shifts, regulations, weather, social sentiment) were identified as causes.

• No experiments or changes in analytics tools have been introduced.

• The measurement method for cancellations has not changed.

Potential Internal Factors Explored:

1. New Feature Releases:

• Some new features were added recently.

• Among them, a virtual try-on tool was launched to help users choose the right product shade. However, it may inadvertently lead to cancellations if users realize post-purchase that they made the wrong selection.

2. Customer Feedback on Cancellation:

• Users mention “product not as expected” as the reason, though this may be a generic response rather than the actual issue.

3. Product Selection Errors:

• Potential confusion around shades or sizes despite the virtual try-on.

4. Sale Timing:

• It’s not linked to upcoming sales—there’s an ongoing sale, which is always pre-announced.

5. Changes in Address Entry:

• No significant recent updates to this process; not seen as a likely cause.

6. Reviews and Perception:

• While the platform is surfacing more positive reviews, customers may still be influenced by negative reviews elsewhere.

• Additionally, the removal of the authenticity badge image (replaced with a text label to improve page load speed) could be reducing trust in product authenticity.

Conclusion:

The likely drivers of the cancellation spike appear to be internal UX and trust-related issues, such as:

• Misuse or misinterpretation of the virtual try-on feature.

• Perceived loss of trust due to removal of visual authenticity cues.

• Discrepancy between what’s promoted (e.g., positive reviews) and what users find externally.

The next step would be to conduct a deep dive into user behavior post-virtual try-on, A/B test the reintroduction of visual authenticity indicators, and gather more qualitative feedback to validate these hypotheses.

⸻

Let me know if you’d like this adapted for a stakeholder deck or interview-ready framework (e.g., root cause analysis format).