Adaptive Ad Load Design for Sponsored Search Markets: Evidence, Theory, and Deployment¶
Fidelity: core mechanism. Paper results, local public-data measurements, and omissions are separate.
Paper information¶
| Field | Details |
|---|---|
| Paper | arXiv 2607.14418 |
| Organization | University of Washington |
| First public release | 2026-07-15 (arXiv v1) |
| Authors' code | The paper did not provide official or author code when checked on 2026-08-24. |
| Adapter / method | adaptive-ad-load |
| Local reproduction code | src/auto_research/reproductions/adaptive_ad_load/ |
Original paper¶
Motivation and method¶
Decide dynamically how many ads to show per search under revenue and conversion constraints. A randomized field experiment estimates the supply curve, then a lightweight policy chooses ad load.
flowchart LR
A["Randomized field experiment"] --> B["Revenue and conversion estimates"] --> C["Adaptive ad count"]Key figure from the paper¶
Figure 14: Overview of the paper's method and key components. From the original paper; copyright remains with the authors.
Core equations¶
Paper results¶
The main online result is revenue +36.80% in a 66-day randomized field experiment. Offline tables and online metrics use private or paper-specific protocols. The private paper protocol is not directly comparable with local MovieLens measurements.
Local reproduction¶
Local comparison: The baseline is the shared transition + content scorer. The treatment adds only adaptive-ad-load on the same users, items, full-catalog candidates, and seed.
MovieLens-100K, 220 users and 360 items, seed 42: NDCG@10 0.0540 → 0.0540 (+0.00%); Hit@10 0.1091 → 0.1091. Validation selects mixture weights only; the test set is not used for tuning.
See the fixed metrics.
Reproduction boundaries¶
The core mechanism runs on MovieLens-100K with the same split, fixed full-catalog candidates, and seed. Private features, the production foundation model, and online traffic are unavailable. The implementation has independent model state and scoring, rather than mapping a paper name to a shared weighted formula. See the adapter's omitted_core_components for omissions. Local changes and paper A/B gains are reported separately.