Practical tips: To improve SMS marketing ROI, these three aspects are the key

 Based on his own practice, the author shares relevant lessons and experience methods about SMS marketing, hoping to inspire you.

I believe that everyone will encounter SMS as a means of operation during the operation process. Although many people say that text messages are a disturbance to users and many people don’t even read text messages (or marketing text messages are directly put into the trash), text messages are still a commonly used operating method. The most important reason is that SMS is cheap and the cost is extremely low! Generally, SMS messages on the market can be sold for a little over 2 cents each, and the ROI can be as high as 152.17, which shows how powerful SMS messages are.

Today I’m going to talk to you about SMS marketing , as well as some experiences and lessons learned from previous SMS marketing.

Let me first give you a brief introduction to the background. We will conduct SMS marketing to users of Pinduoduo stores to further stimulate user consumption. The platform has divided users into several categories according to their attributes, and regular marketing activities can be carried out based on their attributes. You can also analyze user data yourself and do customized SMS marketing for different users. The ROI of SMS marketing can be as high as over 100, and the average ROI is around 40.

Of course, different platforms may have different SMS marketing strategies for different users. Here are some lessons I learned from SMS marketing that I would like to share with you. Everyone is welcome to leave a message below and let’s talk about SMS marketing.

1. Crowdpack is the key

When doing SMS marketing, the audience package is the key. Targeting each type of user with specific SMS marketing can better increase the value of individual users.

Generally speaking, the platform will simply divide users into several categories according to their behaviors (browsing, collecting, purchasing, etc.). As shown in the figure above, the background of Pinduoduo store divides users into those who have not purchased goods in the past 30 days, those who have purchased store goods in the past 365 days, and those who are potentially interested. Generally speaking, based on past SMS marketing data, people who have not purchased store goods in the past 30 days have the highest ROI.

However, although the platform has already labeled the users and prepared some population packages for us, which can be sent directly, the population packages defined by the platform are very general and not precise enough, so we need to further segment the population in order to truly enhance the value of individual users and make SMS marketing a mechanism. (However, in most cases, we do not have the data provided by the platform, so we often need to analyze the data ourselves. The following method is more suitable for most operators.)

Faced with a pile of complex data, how should we segment it, discover the value of each user, and finally enhance the user value through operational means? If you want to deeply explore the value of users, you have to mention the RFM model.

What is the RFM model?

The RFM model is a commonly used model in e-commerce, and is a model for judging user activity and user value.

R represents the time difference of the most recent consumption, which means that the shorter the time difference, the more active the user is and the easier it is to recall. Which one do you think is easier to recall: a user who made a purchase in the past 30 days or a user who made a purchase in the past 90 days?

F represents the frequency of user purchases. The higher the frequency of user purchases, the more active the user is. At the same time, the user's behavior can be inferred based on the frequency of user purchases, and the user can be further recalled.

M represents the consumption amount. The higher the consumption amount, the higher the user value. If we look at the background data and find that user value is generally low, then this means that there is still a lot of room for improvement for this part of users. A user who spends 100 yuan a month and a user who spends 1,000 yuan a month are high-value users that are worth the operator's efforts to maintain. The key is to maintain high-net-worth users while enhancing the value of ordinary users.

According to the RFM model, we can divide users into three groups, so as to better operate according to user behavior.

But what is considered high and what is considered low? Next comes the issue of data settlement, let’s continue. (There are a lot of little Excel tips in this. I finally know the importance of learning Excel well. However, most problems can be solved by using PivotTable.)

How to calculate the RFM model using data?

Taking the backend data of Tmall and Pinduoduo as an example, we can see the user nickname (user's mobile phone number), the date of the user's most recent payment, payment frequency, payment amount and other data. Then we will start processing the data journey.

Calculate R, which is the time difference from the nearest one. Simply subtract two dates to get the most recent consumption difference.

After calculating the most recent consumption difference, we need to determine which R values ​​are called high values ​​and which R values ​​are called low values? There is no single way to determine the high or low R value, it all depends on the type of product and the ultimate purpose that needs to be achieved.

Here, I take the average of the time difference as a judgment criterion (the average R value here is 31 days). A value greater than 31 days is a high R value, and a value less than 31 days is a low R value.

Similarly, it can also be calculated based on the natural usage cycle of the product. If the natural usage cycle is 7 days, then it can be assumed that the R value is greater than 7 days, and the R value is less than 7 days.

After determining the R value, we can use conditional formatting to filter out users who meet the criteria.

Next, we calculate the F value, which is the frequency of consumption by a single user within a certain period of time, that is, the number of times. Here, we define a cycle as one month, and we need to calculate the number of purchases made by a single user in a month.

By using a pivot table, you can calculate the frequency of consumption by a single user within a certain period.

Insert a PivotTable:

Drag the recharge number into the row label, the recharge amount into the column label, set the field to sum, drag the recharge number into the column label, set the field to count, so that the number of consumptions within a certain period of time can be calculated, as shown in the following table:

After the calculation, you can refer to the method of determining the R value above to calculate the F value, which will not be explained here.

In this way, we have determined the data in the RFM model to classify and group users and better carry out refined operations on users.

For example, we can give extra rewards (spiritual or material rewards) to high-value and very important users to stimulate them to continue consuming.

For important development customers, we can recall users regularly. We can conduct regular recalls based on the natural usage cycle of the product or the product consumption. For example, it is easier to recall these users when new products are launched or activities are held.

Every time we compile statistics on the recall data of the population package, SMS click-through rate, ROI and other important data, we constantly adjust the strategy during the SMS marketing process. In fact, this process is also a way to continuously screen out high-quality populations. Through multiple SMS marketing, we can get more accurate population packages, and then through data analysis, we can conduct in-depth operations on the population and enhance the individual value of users.

2. SMS sending time

After briefly talking about using user groups to perform refined user management, let’s talk about the issue of the time of sending SMS messages.

Generally speaking, the best time to send text messages is between 10-11 in the morning, and it is best to send text messages about a week before the end of the event. Sending text messages in the morning is the basis for ensuring that users can see the text messages as much as possible. Sending text messages one week before the end of the event ensures that the event has not ended when users see the text messages, and will not deviate from their psychological expectations.

Here is a small pitfall I would like to share with you. Never send marketing text messages on weekends as the open rate is extremely low. With more choices on weekends, people pay less attention to their mobile phones, resulting in extremely low click-through rates for text messages. Judging from the current SMS data sent, the SMS data on weekends is the worst.

You can try sending text messages on Wednesdays and Thursdays, the click-through rate and payment rate will be higher.

3. SMS copywriting

What kind of marketing text messages are more attractive, and what kind of text message copy will make people more willing to click on them? After all, in this era of text message flooding, if the text messages themselves cannot impress people and make them want to click on them, then basically these text messages will be forever buried in the junk messages.

1. Information that is closely related to users will have a higher opening rate

Everyone cares about things that are closely related to themselves, such as Pinduoduo’s marketing text message: Exclusive discounts for Chaoyang District have been delivered. The xx product you recently browsed is only xx yuan. Click to buy. The opening rate of such text messages is generally quite good. There is an exclusive discount for the user's area, which makes the user feel like they have gotten a bargain.

The second is that the recommended products are related to the user’s recent browsing history. The browsed products indicate that the user is interested in the product and is a potential target user. The last and most important point is that it needs to be supported by marketing activities. The preferential treatment ultimately prompts users to click and place an order.

2. Scenario-related, increasing the credibility of SMS

Creating a specific scenario for users in text messages can, on the one hand, increase users' trust in text messages, and on the other hand, reduce users' aversion to text messages.

What is scene relevance? It means being able to accurately describe the scene the user is currently in. For example, if the temperature is dropping, you can send warm wishes to the user; on the Little New Year, you can send New Year wishes to the user. This can not only facilitate transactions but also make the user feel warm.

In my opinion, SMS marketing is also a means of user management. Based on some historical data of users, users can be stratified and grouped, and users can be operated in a refined manner to enhance the value of individual users.

Between 10,000 users who spend only one yuan per month and 1,000 users who spend 100 yuan per month, which users are more valuable and more worthy of refined operations? I believe we all have the answer in our hearts.

Author: Operation Wang Diary

Source: Operation Wang Diary