Douyin e-commerce data analysis case

Douyin e-commerce , which is in its bonus period, is developing rapidly. This article takes "Yili" as a case study of Douyin e-commerce data analysis, mainly focusing on daily live broadcast operations and large-scale marketing activities. It is recommended for friends who are interested in Douyin e-commerce to read.

Recently, I chatted with many friends who are engaged in e-commerce, and they all said: " The company is going to start Douyin e-commerce, so hurry up and join in while the bonus stage is in place."

You may ask, what is the bonus stage?

There is an important indicator, which is ROI (return on investment) ; in live streaming e-commerce, many people can achieve 1:5, 1:10, or even 1:20.

If you achieve 1:10, the product is sold for 10 yuan, the material cost is 2 yuan, plus the advertising cost of 1 yuan, the profit is 7 yuan, and the profit margin reaches 70%.

This is definitely huge profits generated by dividends.

Recently, I saw the review report data of the "Douyin 818 Good Products Festival" event. The transaction volume of the event increased by 4 times year-on-year, and the order volume increased by 5 times.

We can see that Douyin e-commerce, which is in its bonus period, is developing rapidly . For our data analysis partners, it would be inappropriate not to study how to do data analysis on Douyin.

Let’s take a look together.

1. Douyin e-commerce data analysis scenarios

Here, we choose the brand "Yili" as a case for analysis. In just 4 months, the monthly revenue increased from the initial 170,700 to 24.8554 million in June. Yili's milk is really awesome!

Through the study of Yili's operating methods, there are two main aspects :

The first one is: daily operations ; selling goods through online live broadcast for more than 12 hours every day, and increasing sales over time.

Another is: marketing activities ; selectively participate in relevant activities released by the Douyin platform, such as: 618 promotion, 818 promotion, to achieve rapid revenue growth.

Next, our analysis and research on "Yili" will mainly focus on daily live broadcast operation analysis and large-scale marketing activities analysis.

2. Daily Operation Analysis

In daily live broadcast operations , the most important analysis is to monitor the key indicators of each live broadcast , review the live broadcast, find out the deficiencies, and make optimizations. Next, let’s take a look at the Yili live broadcast review case.

Before we begin, let’s first sort out the user’s behavioral path map from entering the live broadcast room to placing an order.

The reason why Douyin e-commerce is favored by brands is because of its snowball growth logic.

When the live broadcast room is active and users place orders, the platform will recommend the live broadcast room to more new users, and then they will become active again, make purchases, and continue to recommend it , thus achieving a snowball effect.

Based on the user path map, an indicator system was built for analysis. Among them, core indicators such as interaction rate, UV value, and purchase conversion rate directly affect the live broadcast effect.

Based on the important indicators, we organize them into tables to prepare for our live broadcast review.

Based on the above data, we found that the live broadcast on June 9 had 30% more viewers than the live broadcast on June 6, but the sales were only 58% ; so we reviewed June 9, found the problems, and made improvements.

Here we mainly look at three indicators for comparison and find the reasons.

In fact, there are two main reasons why sales cannot be achieved. The first is that the live broadcast is not active and cannot retain people . Here we mainly look at the interaction rate . The second is that people are not interested in the products and are unwilling to buy them. Here we mainly look at the click-through rate and purchase conversion rate.

Comparing the interaction rates , we found that the interaction rate on June 9 was only 24%.

The way to make the live broadcast room active is usually to distribute lucky bags, such as: giving away whole boxes of milk, mobile phones, and sparkling water for users to collect; the condition for participation is to send relevant comments in the live broadcast room, and it will take 10 minutes to collect the prize, which increases the interaction rate and interaction time.

Here we compared the number of times lucky bags were given away during the two live broadcasts. During the 18-hour live broadcast, the number of lucky bags given away on June 9 was 50% less than that on June 6.

The number of lucky bags cannot directly determine the reason for the low interaction rate. Here we will look at how much people like the lucky bags.

The product used in the live broadcast on June 9 was Zhixuan original soy milk , and the product used on June 6 was Anmuxi yogurt from the small black bottle.

We randomly counted the number of people participating in the lucky bag purchase at 9:30 a.m., 2:30 p.m., and 7 p.m., and found that the overall participation rate on June 9 was 50%-80% lower than that on June 6.

Finally, it was concluded that the interaction rate on June 9 was not high . The main reasons were: one, relatively few lucky bags were given away; the other was that people did not like the lucky bags very much. It is recommended to optimize the selection of lucky bags in the future.

Next, we compared the product exposure-click conversion rate and click-purchase conversion rate . For the convenience of display, we optimized the indicator and used the exposure-purchase conversion rate (the product of the exposure-click conversion rate and the click-purchase conversion rate).

Here we compare the TOP3 sales products of the two live broadcasts. The exposure conversion rate of the top 1 on June 9 was 30% of that on June 6, while the top 3 product on June 9 was only 10%. It seems that there is a problem with product selection and it needs to be optimized.

In addition, when watching the products in the live broadcast room on June 9, I found a hidden product. Due to low exposure, it only brought in 14,000 sales, but the exposure-purchase conversion rate reached 6.65%. This product can be used as a key product selection in the future.

3. Marketing Activities Analysis

Before analyzing a marketing campaign, take a look at the campaign process first, and then implement it step by step.

Here we take Yili Douyin 618 event as an example . The goal of the event is mainly to increase sales , but we can analyze the specific target value and see how to set it.

We know that sales = cumulative number of viewers * UV value . In this case, as long as we estimate the UV value and cumulative number of viewers of the 618 live broadcast, we can get the sales forecast.

For UV value prediction, we can use the UV values ​​of the five live broadcasts during the 618 period, calculate the average value, and get 1.44.

To estimate the cumulative number of viewers, we can use the network popularity index of the Douyin platform during last year's Double 11. It was observed that the popularity on Double 11 was 92% higher than that on November 6. Due to the increased promotion efforts of the platform this year, it can be estimated that the popularity on June 18 this year will be at least 100% higher than that on June 13.

It was observed that the live broadcast on June 13 had a cumulative number of 395W viewers , which was quite popular. By June 18, the cumulative number of viewers was estimated to increase by [0.5,1] times, so the cumulative number of viewers was estimated to be [592.5W,790W].

Therefore, we estimate the sales volume to be between 853.2W and 1137.6W.

For activity plans, the main considerations are product selection and user profiling before traffic delivery.

In terms of product selection, we selected the top 10 products with the highest sales in each live broadcast in June, and selected products from large to small based on the three indicators of exposure-to-purchase conversion rate, exposure-to-click conversion rate, and click-to-purchase conversion rate . Here are the 5 products we selected that can be widely exposed during the event.

Before advertising is launched, it is necessary to create a good user profile, which is a profile distribution based on gender and age.

If you have enough time, you can also study which brands your fans are following, or which KOLs they are following and what products they are selling.

1. Activity warm-up

At 9 p.m. on June 17, Yili released a traffic-generating video, which attracted users by announcing a live broadcast in the workshop . In the end, the effect was good, attracting 744,000 people.

2. During the live broadcast

The first ten minutes of a live broadcast are crucial and will determine the exposure of the entire live broadcast. In the first ten minutes of Yili’s live broadcast room, the exposure was increasing at a rate of more than 100% per minute. This is because advertising is targeted at precise users and the volume is increasing rapidly.

After the rapid increase in volume, it is necessary to maintain the activity of the live broadcast room . We can see that lucky bags are constantly given out to encourage user interaction, and 43 mobile phones have been given away , which makes sufficient preparations for conversion.

And on the day of the June 18 live broadcast, at 2 p.m., a video was released that successfully brought in 382,000 traffic.

When explaining products in the live broadcast room, the number of people entering the room will generally decrease. At this time, advertising is needed to increase the number of people, but the ROI of advertising must always be considered and relevant indicators must be monitored.

3. Conduct activity review

Review of goal setting: Overall it is good. The estimated sales volume is [653.2W, 1137.6W], and the actual volume is 1000.83W, which is the average value of the estimated value range. The UV value is 34% lower than expected. This is because the number of people entering the venue from 9pm to 11pm increased significantly, and it was not well handled (which will be explained in detail later).

4. Review of product selection

Among the five products we predicted, three of them entered the TOP5 revenue list of 618 live broadcast, bringing in 101.2W, 95.88W, and 74.56W respectively.

In addition, we looked at the TOP5 sales products on June 18 and found that one product had an exposure – purchase conversion rate of 0.85%, which was 0.08% higher than the TOP1 product, but the traffic was only 1.546 million, less than 30% of the top 1. Adjustments can be made in future live broadcast product selection.

5. Review of the live broadcast room

We found that between 8pm and 10pm, the traffic to the live broadcast room increased, with an average of 22,000 people entering; however, the number of online users was around 6,500, with no significant increase.

Finally, we looked at the official report data of Douyin, and found that the traffic of the entire Douyin platform increased significantly between 8-10 pm.

However, Yili's live broadcast room was unable to accommodate the crowd. At around 12 noon, the average number of people entering the room reached 25,000, and the lucky bags used were mobile phones; while at 8-10 pm, the lucky bags were milk soda.

By looking at the data, the participation rate of lucky bags for mobile phones is 17.2%, while the participation rate of lucky bags for milk soda is 2.25% . Therefore, between 8 and 10 p.m. on the same day, mobile phones with higher participation rates should be used as lucky bags, which can increase activity and conversion.

In fact, when reviewing the situation, there will always be some hindsight. The real live broadcast is changing rapidly . In the end, we saw that the "Yili" live broadcast achieved very good results, reaching a revenue of 10.0083 million.

Author: DataFrog

Source: DataFrog