A Probabilistic Multi-Touch Attribution Model for Online Advertising

Abstract

It is an important problem in computational advertising to study the effects of different advertising channels upon user conversions, as advertisers can use the discoveries to plan or optimize advertising campaigns. In this paper, we propose a novel Probabilistic Multi-Touch Attribution (PMTA) model which takes into account not only which ads have been viewed or clicked by the user but also when each such interaction occurred. Borrowing the techniques from survival analysis, we use the Weibull distribution to describe the observed conversion delay and use the hazard rate of conversion to measure the influence of an ad exposure. It has been shown by extensive experiments on a large realworld dataset that our proposed model is superior to stateof-the-art methods in both conversion prediction and attribution analysis. Furthermore, a surprising research finding obtained from this dataset is that search ads are often not the root cause of final conversions but just the consequence of previously viewed ads.Keywords computational advertising, multi-touch attribution, survival analysis INTRODUCTIONInternet increasingly becomes the leading advertising medium, where online users generate a tremendous amount of feedback information including clicks and conversions. The feedback data reveal the needs/preferences of users, and thus enable online advertising systems to deliver ads to those who are most likely to respond. Nowadays companies spare no effort to attract consumers to visit their websites through various advertising channels, among which display ads and search ads are two dominant types.Recently, researchers from both academia and industry have become more and more interested in analysing the contribution of each advertising channel to user conversion which is known as the "attribution" problem. An accurate attribution model would be of great help for advertisers to interpret the effects of different advertising channels and make informed decisions to optimize their advertising campaigns (e.g., by reallocating advertising budgets). An online advertising campaign is usually launched across multiple channels such as display ads, paid search ads, social media ads, and so on. In most cases, users would have been exposed to the ads from a particular advertising campaign many times before their final conversion, as illustrated in Figure 1. Suppose that a brand X delivers ads through three channels: display, social and paid search: user 1 saw X's display ad at t 1 1 when browsing a webpage, and then saw X's social ad at t 1 2 ; later, she searched for X's products and clicked its paid ad link at t 1 3 ; finally, she made a purchase on X's website at time T 1 . In this case, how should we assess the contribution of those three advertising channels to that user's conversion?A number of attribution models have been proposed and utilized in recent years. Figure 2 shows some representative ones. Most of the existing attribution models widely used in practice are rule-based, and their effectivenesses are limited by thei...

Learning Multi-touch Conversion Attribution with Dual-attention Mechanisms for Online Advertising

“…Another school of MTA modeling is based on the survival theory [12,13,37], which models the conversion event as the predictive goal and estimates the probability for the event occurrence at the specific time while considering the censored data, i.e., the true occurrence time is later than the observation time. Nevertheless, these methodologies focus more on single point prediction and fail to consider the sequential patterns embedded in the user browsing history.…”

“…Since the user conversion is a rare event, we perform negative sampling in data preprocessing. Following [12,37], the sequence preparation and sampling rules are that (i) if the user has multiple conversion events, her action sequence will be split according to the conversion time to guarantee that each sequence has at most one conversion; (ii) we extract the user action sequences with the minimal length of 3 and maximal length of 20 with the sequence duration within 14 days; (iii) all of the user sequences leading to conversion events have been retained and we uniformly sample the sequences without conversions to 20 times of the number of converted sequences.…”

“…The first stage focuses on the conversion estimation performance which has been widely adopted in the conversion attribution task [12,13,37]. Specifically, given the evaluation samples…”

“…More details could be found in the paper. • AMTA is the Additional Multi-touch Attribution model proposed in [12] which was state-of-the-art for this conversion attribution problem. It applies survival analysis to model the conversion estimation and utilizes the hazard rate of conversion at the specific time to model the conversion attribution.…”

“…The attribution credits obtained in these models are heuristically assigned to each user interaction with the advertiser's contents, rather than statistically learned from the data. For example, Ji and Wang [12] proposed that the final conversion would be driven by the additive hazard rate of being converted at the time of each previous touch point, which pre-assumes that the more user exposures, the higher probability of the final user action. As is illustrated in Figure 2 of one real-world dataset used in our experiments, the conversion rate does not necessarily increase w.r.t.…”