Brand Lift test is a widely deployed statistical tool for measuring the effectiveness of online advertisements on brand perception such as ad recall, brand familiarity and favorability. By formulating the problem of interest into a two-sample test on the binomial proportions from the control group (p_0) and the treatment group (p_1), Brand Lift test evaluates ads impact based on the statistical significance of test results. Traditional approaches construct the test statistics based on the absolute difference between the two observed proportions, a.k.a, absolute lift. In this work, we propose a new bootstrap test based on the percentage difference between the two observed proportions, i.e., relative lift. We provide rigorous theoretical guarantees on the asymptotic validity of the proposed relative-lift-based test. Our numerical studies suggest that the relative-lift-based test requires less stringent conditions than the absolute-lift-based test for controlling the type-I error rate. Interestingly, we also prove that the relative-lift-based test is more powerful than the absolute-lift-based test when the alternative is positive (i.e., p1 - p0 > 0), but less powerful when the alternative is negative (i.e., p1 - p0 < 0). The empirical performance of the proposed test is demonstrated by extensive simulation studies, an application to a publicly available A/B testing dataset from advertising, and real datasets collected from the Brand Lift Testing platform at LinkedIn.

In the past decade, the technology industry has adopted online controlled experiments (a.k.a. A/B testing) to guide business decisions. In practice, A/B tests are often implemented with increasing treatment allocation: the new treatment is gradually released to an increasing number of units through a sequence of randomized experiments. In scenarios such as experimenting in a social network setting or in a bipartite online marketplace, interference among units may exist, which can harm the validity of simple inference procedures. In this work, we introduce a widely applicable procedure to test for interference in A/B testing with increasing allocation. Our procedure can be implemented on top of an existing A/B testing platform with a separate flow and does not require a priori a specific interference mechanism. In particular, we introduce two permutation tests that are valid under different assumptions. Firstly, we introduce a general statistical test for interference requiring no additional assumption. Secondly, we introduce a testing procedure that is valid under a time fixed effect assumption. The testing procedure is of very low computational complexity, it is powerful, and it formalizes a heuristic algorithm implemented already in industry. We demonstrate the performance of the proposed testing procedure through simulations on synthetic data. Finally, we discuss one application at LinkedIn, where a screening step is implemented to detect potential interference in all their marketplace experiments with the proposed methods in the paper.

Phased releases are a common strategy in the technology industry for gradually releasing new products or updates through a sequence of A/B tests in which the number of treated units gradually grows until full deployment or deprecation. Performing phased releases in a principled way requires selecting the proportion of units assigned to the new release in a way that balances the risk of an adverse effect with the need to iterate and learn from the experiment rapidly. In this paper, we formalize this problem and propose an algorithm that automatically determines the release percentage at each stage in the schedule, balancing the need to control risk while maximizing ramp-up speed. Our framework models the challenge as a constrained batched bandit problem that ensures that our pre-specified experimental budget is not depleted with high probability. Our proposed algorithm leverages an adaptive Bayesian approach in which the maximal number of units assigned to the treatment is determined by the posterior distribution, ensuring that the probability of depleting the remaining budget is low. Notably, our approach analytically solves the ramp sizes by inverting probability bounds, eliminating the need for challenging rare-event Monte Carlo simulation. It only requires computing means and variances of outcome subsets, making it highly efficient and parallelizable.

Online experimentation, also known as A/B testing, is the gold standard for measuring product impacts and making business decisions in the tech industry. The validity and utility of experiments, however, hinge on unbiasedness and sufficient power. In two-sided online marketplaces, both requirements are called into question. The Bernoulli randomized experiments are biased because treatment units interfere with control units through market competition and violate the "stable unit treatment value assumption"(SUTVA). The experimental power on at least one side of the market is often insufficient because of disparate sample sizes on the two sides. Despite the importance of online marketplaces to the online economy and the crucial role experimentation plays in product development, there lacks an effective and practical solution to the bias and low power problems in marketplace experimentation. In this paper we address this shortcoming by proposing the budget-split design, which is unbiased in any marketplace where buyers have a finite or infinite budget. We show that it is more powerful than all other unbiased designs in the literature. We then provide a generalizable system architecture for deploying this design to online marketplaces. Finally, we confirm the effectiveness of our proposal with empirical performance from experiments run in two real-world online marketplaces. We demonstrate how it achieves over 15x gain in experimental power and removes market competition induced bias, which can be up to 230% the treatment effect size.