LCM-Promo uses a finetuned LLM to allocate promotions
Original title: LCM-Promo: An LLM-Based Uplift Model for Promotion Allocation
In a churn-reactivation A/B test it raised conversion 1% and cut promotion cost 6% versus Segment-EMA.
LCM-Promo is a task-specific Large Commerce Model, a finetuned LLM trained on transactional commerce data to predict user intent and promotion response. It estimates monotonic promotion-response curves from production logs by reading single-token outputs as conversion-threshold probabilities, conditioned on session context such as day and delivery shift. Authors report a production A/B test on churn reactivation against the Segment-EMA baseline.
Unconfirmed: The conversion and cost gains are reported by the authors from one production A/B test.
- The model conditions on session context including day and delivery shift.
- Single-token outputs are interpreted as conversion-threshold probabilities.
- The reported test increased conversion by 1% and reduced promotion cost by 6%.
- Products
- Segment-EMA
- AI models
- LCM-Promo
From the original article
1 min read at huggingface.coWe present LCM-Promo, an LLM-based uplift model for promotion allocation. LCM-Promo is a task-specific Large Commerce Model (LCM): a finetuned LLM trained on transactional commerce data to predict user intent and promotion response.
By Nesta Midavaine, Floris Fok, Paulo Mendes · huggingface.co
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.