A Unified Framework for Marketing Budget Allocation

Abstract

While marketing budget allocation has been studied for decades in traditional business, nowadays online business brings much more challenges due to the dynamic environment and complex decision-making process. In this paper, we present a novel unified framework for marketing budget allocation. By leveraging abundant data, the proposed data-driven approach can help us to overcome the challenges and make more informed decisions. In our approach, a semi-black-box model is built to forecast the dynamic market response and an efficient optimization method is proposed to solve the complex allocation task. First, the response in each market-segment is forecasted by exploring historical data through a semi-black-box model, where the capability of logit demand curve is enhanced by neural networks. The response model reveals relationship between sales and marketing cost. Based on the learned model, budget allocation is then formulated as an optimization problem, and we design efficient algorithms to solve it in both continuous and discrete settings. Several kinds of business constraints are supported in one unified optimization paradigm, including cost upper bound, profit lower bound, or ROI lower bound. The proposed framework is easy to implement and readily to handle large-scale problems. It has been successfully applied to many scenarios in Alibaba Group. The results of both offline experiments and online A/B testing demonstrate its effectiveness.

Adversarial Learning for Incentive Optimization in Mobile Payment Marketing

“…To allocate the budget in an optimal way, a two-stage solution [13,22] is a common choice. First, a response model 𝑦 = 𝑓 (𝑥, 𝑡) estimates the users' MPPs for each user described by 𝑥 and each price 𝑡.…”

“…However, as a black-box model, there are some gaps between the prediction and decision-making of deep neural networks [2]. Therefore, [13,22] propose a semi-black box model that extends the logarithmic demand curve through neural network and graph learning to solve this problem. Allocation Optimization.…”

E-Business: área considerável de oportunidades de empreendedorismo / E-Business: a considerable area for entrepreneurial opportunities

“…O processo de DNP consome regularmente uma fatia importante dos recursos do empreendedor e da eficiente alocação desses recursos depende boa parte do seu sucesso. A eficiência dessa alocação pode ser substancialmente melhorada com a utilização de data-driven methods (Payyappalli & Zhuang, 2019;Zhao et al, 2019).…”

Free Lunch! Retrospective Uplift Modeling for Dynamic Promotions Recommendation within ROI Constraints

“…CAT E Y (x) is defined as the conditional increment in probability of completing the purchase Pr (Y = 1) caused by the incentive, given the customer characteristics x. CATE estimation can be conducted via various meta-learning techniques [8], such as the two-models approach [21], transformed outcome [10], uplift trees [23], and other meta-learners [14]. Some previous work aimed to address allocation of promotions by estimating the marginal effect [26], incorporating net value optimization within the meta-learner [27] and modeling the optimization as a min-flow problem [18]. Such estimations might be unreliable [4,16], however, due to high variance in the training dataset, which is heavily dependant on a majority of non-transactional visits having no intent to purchase.…”

“…In addition, offline solutions might be biased towards historical data and require dynamic calibration [28] according to long-term changes and seasonality trends, which are particularly common in the travel industry [2] and in promotional campaigns [12]. Dynamic adaptive strategy adjustments is a common practice in cases of budget constraints and were found effective in ads management [17], influence maximization [7,24], marketing budget allocation [1,15,26] and discount personalization [6].…”