E-commerce multichannel direct messaging 2021-2023

### About

This dataset contains multi-channel messages of medium sized online store for 2 years.

Channels: email, web push, mobile push, SMS.

Campaign types: bulk, triggers, transactional.

Notice: this dataset contains `messages-demo.csv` file limited to 10M messages. The full file with 721M messages is available here: [messages.csv](https://data.rees46.com/datasets/direct-messaging/messages.csv.gz) (21.5Gb).

Data collected by [REES46 CDP](https://rees46.com/) project.

### Files description

**Here is a brief description of files. You can find the detailed description of every property in [this notebook](https://www.kaggle.com/mkechinov/direct-messaging-campaigns-dataset-overview).**

#### holidays.csv

Bulk campaigns usually set before holidays and sale outs. This file contains not full list of Russian holidays and commerce activities so you can see how bulk campaigns are related to these dates.

#### campaigns.csv

All messages are related to some kind of campaigns:

1. Bulk campaigns are sent for sale outs and before holidays to stimulate sales and bring back customers.
2. Trigger messages (like abandoned cart) are sent automatically based on user's behavior. More users visited website/app – more trigger messages are sent.
3. Transactional messages are used for some kind of information delivery process: bonuses added, order delivery status changed, etc.

Keep in mind: `campaign_id` is unique only for the specific `campaign_type`. Two campaigns with different `campaign_type` can have the same `campaign_id`. So the unique campaign identifier is `campaign_type + campaign_id`.

Additional properties are added to every campaign to describe its meaning (topic) or subject parameters (emoji, call to action, etc).

#### messages.csv or messages-demo.csv

`Messages` table contains a list of all messages sent with its statuses and meta info:

1. Campaign
2. Channel
3. Type
4. Opened (when)
5. Clicked (when)
6. Purchase
7. Etc. See attached dataset for detailed info of every property

#### client_first_purchase_date.csv

The file has 2 columns:

1. Client ID
2. Date of the first purchase ever

### More datasets

Checkout another datasets:

1. https://www.kaggle.com/mkechinov/ecommerce-behavior-data-from-multi-category-store
2. https://www.kaggle.com/mkechinov/ecommerce-purchase-history-from-electronics-store
3. https://www.kaggle.com/mkechinov/ecommerce-events-history-in-cosmetics-shop
4. https://www.kaggle.com/mkechinov/ecommerce-purchase-history-from-jewelry-store
5. https://www.kaggle.com/mkechinov/ecommerce-events-history-in-electronics-store
6. https://www.kaggle.com/datasets/mkechinov/ecommerce-purchase-history-from-jewelry-store
7. [NEW] https://www.kaggle.com/datasets/mkechinov/direct-messaging - you're reading it right now

### Many thanks

Thanks to [REES46 Marketing Platform](https://rees46.com) for this dataset.

### Using datasets in your works, books, education materials

You can use this dataset for free. Just mention the source of it: link to this page and link to [REES46 Marketing Platform](https://rees46.com).

{'About': '本数据集收录了为期两年的中型在线商店的多渠道信息，共计两年。', 'Channels': '渠道包括电子邮件、网页推送、移动推送和短信。', 'Campaign types': '营销活动类型分为批量发送、触发式发送和事务性发送。', 'Notice': '请注意，本数据集包含的 `messages-demo.csv` 文件仅限于 10M 条消息。完整文件（包含 721M 条消息）可通过以下链接获取：[messages.csv](https://data.rees46.com/datasets/direct-messaging/messages.csv.gz)（21.5Gb）。', 'Data collected by': '数据由 [REES46 CDP](https://rees46.com/) 项目收集。', 'Files description': '以下是对文件内容的简要描述。您可以在[本笔记本](https://www.kaggle.com/mkechinov/direct-messaging-campaigns-dataset-overview)中找到每个属性的详细描述。', 'holidays.csv': '批量营销活动通常在节假日和销售旺季前设置。此文件包含部分俄罗斯节假日和商业活动列表，以便您了解批量营销活动与这些日期的关联。', 'campaigns.csv': '所有消息都与某种营销活动相关联：
1. 批量营销活动旨在促销和吸引顾客，通常在节假日和销售旺季前发送。
2. 触发式消息（如购物车 abandonment）根据用户的自动行为发送。网站/应用程序访问的用户越多，发送的触发式消息就越多。
3. 事务性消息用于信息传递过程：如添加奖励、订单配送状态变更等。
请注意：`campaign_id` 在特定 `campaign_type` 下是唯一的。不同 `campaign_type` 的两个活动可以具有相同的 `campaign_id`。因此，唯一的营销活动标识符是 `campaign_type + campaign_id`。
每个营销活动还添加了额外的属性，以描述其含义（主题）或主题参数（表情符号、号召性用语等）。', 'messages.csv or messages-demo.csv': '`Messages` 表包含所有发送的消息列表及其状态和元信息：
1. 营销活动
2. 渠道
3. 类型
4. 打开时间（何时打开）
5. 点击时间（何时点击）
6. 购买
7. 等等。请参阅附加的数据集以获取每个属性的详细信息。', 'client_first_purchase_date.csv': '该文件包含两列：
1. 客户 ID
2. 首次购买日期', 'More datasets': '请查阅其他数据集：
1. [https://www.kaggle.com/mkechinov/ecommerce-behavior-data-from-multi-category-store](https://www.kaggle.com/mkechinov/ecommerce-behavior-data-from-multi-category-store)
2. [https://www.kaggle.com/mkechinov/ecommerce-purchase-history-from-electronics-store](https://www.kaggle.com/mkechinov/ecommerce-purchase-history-from-electronics-store)
3. [https://www.kaggle.com/mkechinov/ecommerce-events-history-in-cosmetics-shop](https://www.kaggle.com/mkechinov/ecommerce-events-history-in-cosmetics-shop)
4. [https://www.kaggle.com/mkechinov/ecommerce-purchase-history-from-jewelry-store](https://www.kaggle.com/mkechinov/ecommerce-purchase-history-from-jewelry-store)
5. [https://www.kaggle.com/mkechinov/ecommerce-events-history-in-electronics-store](https://www.kaggle.com/mkechinov/ecommerce-events-history-in-electronics-store)
6. [https://www.kaggle.com/datasets/mkechinov/ecommerce-purchase-history-from-jewelry-store](https://www.kaggle.com/datasets/mkechinov/ecommerce-purchase-history-from-jewelry-store)
7. [NEW] [https://www.kaggle.com/datasets/mkechinov/direct-messaging](https://www.kaggle.com/datasets/mkechinov/direct-messaging) - 您现在正在阅读它', 'Many thanks': '感谢 [REES46 Marketing Platform](https://rees46.com) 提供此数据集。', 'Using datasets in your works, books, education materials': '您可免费使用此数据集。只需提及数据来源：本页面的链接以及 [REES46 Marketing Platform](https://rees46.com) 的链接。'}

搜集汇总

背景与挑战

背景概述

该数据集包含2021-2023年间一个中等规模在线商店的多渠道直接消息数据，涵盖电子邮件、网页推送、移动推送和短信等多种渠道，以及批量、触发和交易性三种营销活动类型。数据集提供了消息的详细状态和元信息，以及相关的假日和商业活动信息，适用于电子商务营销分析和用户行为研究。