Forum Posts

Mahfuj Alom
Jun 19, 2022
In General Discussions
Editor's introduction: How can operators combine data analysis to find marketing-sensitive groups, improve reach and conversion effects, and reduce bulk sms service marketing costs? Take a look at the author's case study. In this article, the author combines the Amazon SageMaker Canvas product to conduct marketing scene modeling practice, let's take a look. In the post-Internet era, with the rising cost of marketing, how to accurately find marketing-sensitive groups from the stock crowd to reach them, and then improve ROI has always been an important topic in business. Such business scenario requirements also extend to the inspection of data bulk sms service analysts' capabilities. For example, there is such a high-frequency business interview question: If Ele.me intends to accurately issue coupons to users, how to predict which users will use it? Discuss the problem at the business level. The reason for predicting the people who will use coupons is to maximize the marketing output bulk sms service under the premise of limited costs. The key point is to find the people who are really moved by marketing, namely Marketing sensitive people. 1. The theory of marketing gain model In the field of digital marketing, there is a classic marketing gain model uplift modeling, which can help us achieve this goal. The uplift model divides users into four categories according to the two dimensions of marketing intervention (such as coupons) and intervention results (whether to buy or not): Marketing-sensitive groups Persuadables: Don't buy if you don't send a coupon, buy if you send a bulk sms service coupon; Naturally converted people Sure things: whether they send coupons or not, they will buy; Lost crowd Lost causes: No purchase regardless of whether coupons are sent; Anti-advertising crowd Sleeping Dogs: If you don't send a coupon, you will buy it, but if you send a coupon, you won't buy it. In order to maximize the efficiency of marketing bulk sms service conversion, our idea is to identify the marketing sensitive groups (Persuadables) and issue coupons to them. Before discussing how to find marketing sensitive people, let’s take a look at how to define this group of people from the data level? Because data prediction is based on probabilistic thinking, the previous definition of the crowd can be replaced with probability: when coupons are issued, the probability of purchase is high; if coupons are not issued, the probability of purchase is bulk sms service small. Further, the expected income can be calculated when the bonds are issued and when the bonds are not issued, and the income difference can be obtained. This difference in return is the "gain",
Uplift Model for Marketing Bulk Sms Service content media
0
0
3
M
Mahfuj Alom

Mahfuj Alom

More actions