Shulou(Shulou.com)06/01 Report--

What this article shares to you is about the scene and function around Target Group in SAP Marketing Cloud. The editor thinks it is very practical, so I share it with you to learn. I hope you can get something after reading this article.

I'll introduce you to some of the scenarios and features around the concept of Target Group in SAP Marketing Cloud.

(1) Segmentation of users based on tags: Segmentation Modeling

In Segmentation Modeling, users can be subdivided by various tags, such as product ID, interaction mode, region, gender, birthday, name and so on. The data are presented graphically and other forms after statistics. The way to subdivide users is simple and convenient, and it can be easily realized by checking or clicking.

(2) Forecast Studio Predictive Studio

With Predictive Studio, business analysts can create forecasting models. The prediction model uses algorithms and historical data calculations to provide scores for future customer behavior analysis.

Scenario example: Emma, the business person in charge of product A marketing, wants to achieve 200 sales of the product through a mobile marketing campaign.

An overview of the process for creating and using predictive models:

(1) in Predictive Studio, create a prediction model, select a prediction scenario, and define the details required for the scene.

(2) create one or more model fitting for the prediction model.

(3) using historical data to train the model.

(4) check the quality of model fitting and select the best model to merge and activate the prediction model.

(5) the best model fitting can be used to calculate the prediction score.

(6) in Segmentation, create a target group based on the active prediction model.

(7) in Campaigns, run a marketing campaign for target group.

(8) in Predictive Studio, measure the success of a marketing campaign to see a way to optimize the activity model in the future.

The following are the specific steps.

Create a prediction model

The Predictive Studio page shows the existing prediction model, and we can choose to create another one.

In the scenario mentioned earlier, the prediction scenario Scenario should choose Consumer Buying Propensity.

Make the following settings:

Training Set: target group of the training set (no more than 1 million members)

Target Object: product A

Target Variable: purchase

Time Frame for Analysis: specifies the analysis period for the training set

Number of Responses: the number of members who purchased product An in the training set (some people in the training set must have purchased product A)

Number of Members: number of members in the training set

Applicable Scope: specify the area where the training set is valid

After setting, select save.

Create a model fit

First of all, the internal training model is fitted. In the Model Fits section, we can create a model fit.

On the model fitting page, we select the Predictors associated with the model, and then click Start Model Training to start the training.

At the end of the model training, the following results are displayed, and the meaning of the display is as follows:

Predictive Power: belongs to [0jue 1], indicating the fitting quality of the model, the bigger the better.

Predictive Confidence: belongs to [0J1], indicating the confidence level of prediction, the bigger the better. We think that more than 0.95 is reliable.

Initial No. Of Predictors: predicts the number of entries in its list that works.

No. Of Selected Predictors: the number of predictors selected.

No. Of Kept Predictors: the number of active predictors retained.

The bar chart represents the percentage of each active predictor that works.

Upload external training model fitting

In addition to using the standard predictive analysis model provided by SAP, we can also upload external training model fitting. When creating a new model fit, you need to select Logistic Regression.

At the bottom of the simulation fitting details page, click Import Model to import. Note: only files in xml format can be imported.

If the model is imported successfully, the Lorentz curve is calculated using the training set and the curve is displayed in the Predictive Model chart.

Select the best model fitting

We can create a few more Model Fit and choose the one that fits best. The criteria for measuring the fitting effect are as follows:

Quality Coefficient: that is, the mass coefficient (also known as the Gini coefficient), with a value of [- 1], which is proportional to the area between the random line and the model curve, indicating the quality of the model fitting.

Lorentz curve: the following figure visualizes the fitting quality.

Select the best fitting Model Fit for Activate, and the activated model will display Active.

Apply the prediction model to the user segmentation of marketing activities

You can subdivide users by double-clicking a country based on the graphical interface:

Select Buying Propensity, find the model we set up in advance, fill in the expected order number 200 in the Predicted Expected Responses column, and Selected Contacts will display the target group size. Select Keep and determine Selected Contacts as the target group.

At this point, we can create a Target Group based on the results we have broken out. Marketer Emma can use this Target Group to carry out intelligent and efficient precision marketing activities.

You can see the information of Target Group. Click Release before it can be used in Campaign.

Score Builder Scores Builder

Set up Score standards to supplement user portraits as the basis for user segmentation.

The Score Builder home page displays the existing Scores, which can only be viewed and cannot be modified.

Click to view Score details to view the Rule Model used by Score, the folder you belong to in Segmentation Modeling, and the applications that can use the Score.

Click Create Score on the home page of Score Builder to create a new Score.

Click the plus sign to create a Rule Model. A Score can have multiple Rule Model, which supports a combination of different Rule settings. Rule is in the form of if then, and different rules are connected with "and" or "or". Compared with simply using tags to segment the audience in Segmentation Modeling, Rule Model provides the possibility for a variety of attributes to be measured comprehensively according to different weights.

After filling in other information such as Target Group and Time Frame, save and activate. At this point, the creation of Score is complete.

A Score Rule named Best Email Sending Time automatically counts the number of users with a status of Active for the entire customer in each time period.

We can select the time period with the largest number of active users, click Keep, and subdivide the audience again. The process for creating a Target Group is the same as above.

Subdivide according to user behavior-- trigger-based Marketing campaign Trigger-Based Campaigns

In addition to the methods mentioned above, we can also subdivide users according to their behavior. Once the user makes these behaviors, they join our Target Group. This type of Target Group automatically triggers the opening of the marketing campaign.

This kind of behavior includes: Abandoned shopping cart/App installed/Email opened/Email not Opened/Click through/No click through and so on.

For this kind of automatically triggered marketing activities, when creating a Campaign, you need to set the Trigger Type in the User Interaction, that is, to set the user behavior that triggers the marketing campaign. In this way, when the user makes these actions, the system will react automatically. For example, when users abandon the shopping cart, send an email to the user to remind them of the goods in the shopping cart, which can improve the profitability of the online mall to a certain extent.

Determine the audience of marketing activities-Sentiment Engagement according to the analysis of emotional interaction

When analyzing the user's emotional interaction in the Sentiment Engagement function, we can set the Target Group according to the user's behavior. First create a new Target Group, and then add individual users to it.

Based on the filter criteria showing eligible emotional interactions, we can check the user's Post and import the previously created Target Group.

Because you need to add users one by one, this method is suitable for a small range of marketing activities.

Determine the audience of the marketing campaign according to the customer journey analysis-Customer Journey Insight

Due to data problems, the following two pictures are screenshots of the official website. The following shows the top 80% return trips through 8 contact points over a period of time.

Based on the customer journey data that has been filtered and analyzed, select the desired part to build the Target Group.

These are the scenes and functions around Target Group in SAP Marketing Cloud. The editor believes that there are some knowledge points that we may see or use in our daily work. I hope you can learn more from this article. For more details, please follow the industry information channel.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.