# Novelty and Primacy: A Long-Term Estimator for Online Experiments

arXiv: https://arxiv.org/abs/2102.12893 · HTML: https://mlsanity.com/paper/2102.12893/

## Key facts

- Cited by 5 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 18 Feb 2021.

## Built on by (most cited first, 5 of 5)

- [Not Just Novelty: A Longitudinal Study on Utility and Customization of an AI Workflow](https://mlsanity.com/paper/2402.09894/) (2024)
- [Facilitating Longitudinal Interaction Studies of AI Systems](https://mlsanity.com/paper/2508.10252/) (2025)
- [Validation of massively-parallel adaptive testing using dynamic control matching](https://mlsanity.com/paper/2305.01334/) (2023)
- [Causal Estimation of User Learning in Personalized Systems](https://mlsanity.com/paper/2306.00485/) (2023)
- [Efficient Multi-Cohort Inference for Long-Term Effects and Lifetime Value in A/B Testing with User Learning](https://mlsanity.com/paper/2604.20777/) (2026)

## Abstract (arXiv)

Online experiments are the gold standard for evaluating impact on user experience and accelerating innovation in software. However, since experiments are typically limited in duration, observed treatment effects are not always permanently stable, sometimes revealing increasing or decreasing patterns over time. There are multiple causes for a treatment effect to change over time. In this paper, we focus on a particular cause, user-learning, which is primarily associated with novelty or primacy. Novelty describes the desire to use new technology that tends to diminish over time. Primacy describes the growing engagement with technology as a result of adoption of the innovation. User-learning estimation is critical because it holds experimentation responsible for trustworthiness, empowers organizations to make better decisions by providing a long-term view of expected impact, and prevents user dissatisfaction. In this paper, we propose an observational approach, based on difference-in-differences technique to estimate user-learning at scale. We use this approach to test and estimate user-learning in many experiments at Microsoft. We compare our approach with the existing experimental method to show its benefits in terms of ease of use and higher statistical power, and to discuss its limitation in presence of other forms of treatment interaction with time.