This article is compiled according to the live sharing of “Business Growth Plan of brand retail Industry” by Han Rui, head of brand retail industry department of Shence Data. The main contents of this paper are as follows:

  • Shence data enables the overall blueprint of the brand retail industry

  • Attribution of realizable value of public domain

  • Private domain population asset health diagnosis

  • The health of the private sector has improved

1. Shence data enables the overall blueprint of the brand retail industry

In recent years, the interaction points between consumers and brands have become increasingly diversified, fragmented and omni-channel, and the consumption scene has become increasingly rich.

Pictured above, social platform, electric business platform and offline channels has become a new force three for brand retail industry, and in this circle, scenes and life scenes of the current domestic consumption has and inseparable, which has a different life scene, a consumer of the day to be able to inspire the consumption of different scenarios. In this context, Shence Data is committed to creating a thriving private ecosystem for brand retailers, enabling the full range of realizations, and thus enhancing the long-term value of brands.

Shenze creates a thriving private domain in two ways: the first is the technical side and the second is the business side. On the technical side, we will first help the brand build digital infrastructure in private domain. Shence has a relatively mature solution — CDP solution, which can help brands to realize the data closed-loop of “acquisition, management, general calculation and application”. Simply speaking, it can realize the five-step closed-loop of data collection, governance, opening, calculation and application.

For the business side, in terms of the current status of the brand, both online private domain and offline private domain may become the private domain of the brand. Online, a large number of brands have their own online contacts, in the form of small programs, official accounts, H5 official website as carriers; Offline, a large number of brands have their own stores, which constitute the premise of the brand online and offline private domain linkage. For the offline private domain Of the brand, Shence helps the brand realize the integration Of Online and offline channels from data to crowd to operation through OMO program, namely online-merge-of Fline. As shown in the figure above, Shence helps brands integrate online and offline private domains through business growth and OMO.

In addition, in the current domestic Internet pattern, in addition to private domain, public domain still accounts for a large proportion. One part is the online public domain represented by e-commerce platform, media platform and social platform. One part is the offline public domain represented by supermarket, department store and convenience store. When the private domain of the brand can fully realize mutual empowerment with the online and offline domain of the public domain, the private domain can be made bigger and the public domain can be realized at the same time.

Shence data provides a complete solution for the private sector business growth of brand retail industry (as shown in the figure below), which includes three parts: attribution of public domain realization value, diagnosis of private domain population’s asset health, and improvement of private domain population’s asset health. The specific contents will be explained below.

Within an enterprise, the private business department should cooperate with the public business department, otherwise it is difficult for the private business department to get the attention and budget support of the company’s top management. The private sector of the brand retail industry needs to “flag up” the value of the brand as a private sector and the empowerment of the public sector.

2. Attribution of public domain realization value

Nowadays, the private sector of the brand retail industry has the following pain points:

  • Private domain mall has disadvantages in GMV contribution in the short term

  • How to prove the value of a private domain business if operating it does not bring about a qualitative change in sales in the short term?

The way out of these pain points is that some core behavior of consumers can connect the public domain with the private domain and prove the value of the prosperity of the private domain to the realization of the public domain. Through the powerful attribution model, the divine policy data can truly restore the prior behavior path of consumer core behavior, so as to achieve attribution traceability.

Supplement knowledge



What is attribution analysis?

  • The completion of an action is the core action

  • According to some sort of calculation rule

  • Retrace the prior behavior forward

  • Custom set the preorder behavior window period

  • Distribution of credit for this goal

What problems can attribution analysis solve?

  • Clearly restore the antecedent behavior path of core behavior

  • Truly reflect the contribution of each antecedent behavior to the core behavior

Case: The public domain realization value attribution of a top consumer electronics brand taking “registered products” as target behavior

The top consumer electronics brand hopes consumers will become registered users after buying its products on Tmall, JD.com and various offline channels.

Using shenpolicy, the brand found that all registered product users will have the following behavior characteristics:

First, user A has never visited the mini program of the brand before registering the product on the mini program, which means that user A does not belong to the “private domain crowd”.

Second, users B and C have already browsed the product category page, details page, articles and watched live broadcasts in the private domain applet. On June 18, users habitually placed orders on Tmall and JINGdong, probably because these e-commerce platforms were very large and did a good job in logistics and delivery services. Users still paid on e-commerce platforms. But as long as he can come back to register the product, he will have the opportunity to find that he has been trained in the private domain loyalty before placing orders on JINGdong and Tmall, but only the last step of placing orders, went to the e-commerce platform. This logic is very important for brands, we can use the following figure for a logical deduction.

Suppose that there are 5 million people in the private domain and 160 million people in the public domain. During the promotion period of June 18 and November 11, 40,000 people purchased the private domain and 2.4 million people purchased the public domain. We guess that today’s consumers are more accustomed to placing orders on e-commerce platforms, and most users belong to the above two categories of B and C, only completing the cultivation of loyalty in the small program. Then how to verify this logic?

Suppose that through brand operation, 200,000 of the 2.4 million people who purchased products in the public domain return to the mini program to register products, and the remaining 2.2 million do not register products. According to the attribution calculation model, 30,000 of these 200,000 people have been trained privately before placing orders on e-commerce platforms such as Tmall and JD.com. So, in fact, out of the 5 million people in the private domain, there are two groups of people, one is the 40,000 people who buy directly in the private domain, and the other is the 30,000 people who buy directly in the public domain.

To sum up, private domain can be used to enable global realization, which affects the actual value of 70,000 people, rather than the initial simple 40,000 people. This logic deduces, from the perspective of data, to help the private business department of the enterprise prove the value of the team to the company’s senior management.

Iii. Health diagnosis of private domain population assets

Health diagnosis is the first and crucial step in the asset operation of private population. Using AIPL consumer journey to depict the health of private population has basically become the mainstream consensus of brand retail industry, and brand retail enterprises hope to embrace AIPL. After the crowd precipitates to the private domain, each enterprise needs to make a personalized definition of the crowd assets, and the broad AIPL definition cannot meet the actual needs of the enterprise. With the definition, the enterprise also needs to master the quantity and quality of the assets of the private domain crowd at any time, and timely find the problem, and the right remedy.

In each period, Divine Strategy will help enterprises find 4 indicators of AIPL population in each stage and 3 indicators of vertical transition, which are used to measure the asset health of private population, as shown in the figure below.

Through this coordinate system, the population asset growth in different time periods is compared horizontally, the population asset transition is observed vertically, and the health of the brand’s population asset in a certain time period is measured. As can be seen from the figure, the cognitive population of brand A doubled from January to February, from 100,000 to 200,000, while the longitudinal transition of cognitive population decreased from Month to month. Similarly, the transition rate of interested population to buying population also decreased from month to month, while the longitudinal transition of buying population remained unchanged from January. It can be seen that brand A should focus on the transition from cognitive group to interested group and from interested group to buying group in March, and explore what problems existed before and how to optimize.

Shence first realizes One ID identification of private domain consumers through full-end data collection, governance, opening and calculation. Such as consumers in the brand on the precipitation of some historical data, such as CRM data, order information, membership information, marketing & customer service data and existing labels, free platform data, such as source, through registration, add to cart, pay orders, user information, such as electricity order data, such as taobao, jingdong, Tmall, praise, etc., and other third-party data. Shence will get through these all-end data to realize the One ID of consumers in the private domain. Then, Shence will help enterprises to customize the generation of AIPL crowd label, and then diagnose the health degree of crowd assets.

Iv. The health of private assets is improved

1. Improve cognitive population: optimize the delivery to improve the quality and quantity of cognitive population

Brands need to accurately identify private target groups, guide channel launch, and lean iteration. First, according to different private domain business objectives, excavate target group portraits and accurately develop channels. Second, the whole channel of automatic monitoring, every penny spent clearly. Third, from the pull of new quantity and quality evaluation of channel delivery effect, optimize channel delivery. Shence also through these three steps, to help brand retail enterprises to solve the pain point.

First of all, Shence helps enterprises establish target group portraits according to personalized business objectives, develop accurate channels, and prepare for launch. As shown in the figure, it is assumed that the users who have completed the purchase conversion on the day after registering in the private domain and have re-purchase behavior within 10 days are the most important users for the enterprise at present. Then, such users can be regarded as the new target and the group characteristics of the target users can be analyzed through the magical strategy. Through drilling, the study found the group of 18-45 years old male white-collar professionals, first-tier cities, and they browse the topic more distribution in the NBA, tourism, decoration, food, clothing shoes and hats, and other fields, and so will extract such labels, help the brand well suited to screen out such user community channels and advertising, sufficient preparation for the launch strategy.

Secondly, Shenpolicy helps enterprises to achieve whole-channel monitoring and tracking, so that every investment has a trace to follow. At present, Shence has covered some mainstream promotion platforms, including SEM/SEO promotion, small program fission, third-party marketing activities, app stores and other promotion channels. At the same time, Shence has automatic channel analysis and comparison components, brands can effortlessly obtain the transformation effect of one click. For example, an enterprise advertises on both Toutiao and Baidu, and finds that although toutiao has a small number of users, the next-day retention and per-capita payment capacity of those users are higher than Baidu. These are very important findings.

Finally, Smart strategy can help enterprises analyze the private performance of cognitive groups from various delivery channels, truly evaluate the delivery ROI and optimize the delivery portfolio. After completing real-time omni-channel monitoring and tracking, Shence helps enterprises analyze the follow-up performance of cognitive groups in each channel, so that they can truly evaluate the ROI. For example, some channels to the flow will have a more active registration behavior, but after registration only get coupons do not place orders, not more interaction with the brand. Shence can objectively assess ROI by tracking user behavior in private domain traffic through the full life cycle, and also help brands optimize their placement mix.

Case: A consumer finance customer made precise channel investment by identifying target consumer groups

The customer is a leading consumer finance enterprise, with more than 15 provinces (cities) marketing centers, business throughout the country, providing consumers with all-round financial services. However, in the process of development, there are two pain points:

In the process of digital transformation, online data has not been combined with business data, so there is a lack of understanding of the portrait characteristics of online crowd

Advertising was carried out in the office buildings in the core business district to attract the target group of white-collar workers. However, with only PU and UV data, it was difficult to evaluate the channel matching and could not measure the actual ROI

Through the analysis of the multi-dimensional data, We found that:

  • The borrowers were portrayed as young students with low incomes or new employees, not white-collar workers

  • Daily to hundreds of thousands of office advertising, the actual lending conversion rate is almost 0, capital registration conversion is minimal

Therefore, we decided to target the young and low-income groups, and adjusted the marketing strategy, and changed the marketing channel from office building to short video platform. As a result, the conversion rate of new channel registration increased by 120%, the conversion rate of loan increased by 38%, and the delivery cost decreased by 50%.

2. Improve the interest group: improve the transition from cognitive group to interested group

The brand hopes to convert more cognitive groups into interested groups and carry out fine operation for interested groups. The difficulty is:

  • How to create an attractive online touch point, accurately push the content that users are interested in, and convert more cognitive groups to interested groups

  • Do interested groups need to be subdivided? If so, how to find the characteristics of shallow and deep interested groups and transition more shallow interested groups to deep interested groups

In view of the above difficulties, Shence helps enterprises create online contact points that are attractive to cognitive groups through personalized recommendation programs, and stimulate interest in interacting with brands. For example, in a private domain small program, there are both personalized product recommendation, personalized classified goods ah, there are personalized buyer show, and the theme of good thing recommendation, new product recommendation, popular popular style recommendation and so on. When a consumer enters this private domain, the content he sees is personalized recommended products, buyers show, categories, popular styles and new products generated according to his portrait, which are quite attractive to him. In terms of the implementation method, the smart data circle the crowd according to the behavior and attribute tag, and match the promotable goods, calculate the potential click rate ranking, and then display.

Case: A top consumer electronics brand realized personalized recommendation through crowd portrait and automated operation, thus creating more than a thousand consumers

In the brand’s public number, there is a male and a female two consumers are concerned about the public number, but their fact tag is different, male consumers are members of the brand, from the natural flow; Female consumers are not members of the brand and come from Paid traffic. At the same time, the behavior labels of the two consumers are also different. Male consumers have bought a product of the brand, are interested in the article about “vacuum cleaner” in the official account, and often visit the official account at 8 PM, and even initiate two searches by themselves. Female consumers have not purchased the products of this brand, are interested in the articles about “hair dryer” in the official account, and often visit the official account at 9 am. They have never launched their own search, and the Feed reading times is 3. Therefore, according to the crowd portrait of the two, the home page will recommend vacuum cleaner, curling iron and other products to achieve more than 1,000 people.

On the whole, it is necessary to transition the cognitive group to the interested group, but the probability of the transition from the interested group with different characteristics to the buying group may be different. The order attribution can find the characteristics of the interested group that has the greatest influence on the purchase transformation.

Case study: A top consumer electronics brand finds the “magic number” of people interested in jumping

First, the core forward event that browsed the goods page to purchase is found through order attribution. After that, the 7-day retention rate, purchase conversion rate, registered product conversion rate and so on can be found through the mini-program to browse the product details page for 5 times as the transition node of interested people. Then, in the business scenario of the brand, browsing the product details page ≥5 times can be defined as deep interest, otherwise defined as shallow interest. Finally, further explore the characteristics of shallow and deep interest groups, find the overall user transition mode, find the business can execute the transition scheme.

When a brand finds the magic number, it often finds the most important factor influencing the conversion of purchases.

Case study: a top consumer electronics brand finds the transition pattern of interested people

As shown in the figure below, consumer A browses the product page for 5 times during the pre-promotion period, thus switching from shallow interest to deep interest. The increase of searching and browsing articles is the core factor for the transition. Then, the operation idea changed from A to multi-initiated search and multi-browse articles, from unexecutable to easy to execute, from untouchable to touchable. Finally, the enterprise extends the case methodology to the whole platform consumers, and finds that watching live broadcasts and browsing articles more can realize the overall transition of shallow interested groups.

3. Promote the buying crowd: promote fine operation and purchase transformation

In the 618, Double 11 and other promotion period, how to improve the purchase conversion of interested groups at all stages, and improve the operation efficiency, is the top priority of the brand. At ordinary times, through diversified operation means, we strive to manage the “grass planting period” well. In the first 20 or 30 days of promotion, we continue to push content to activate users and enter the “warm-up period”. In the first 7-10 days of promotion, real-time messages stimulate users, reduce users’ hesitation and enter the “harvesting period”. The so-called “real-time message stimulation” means to make differentiated coupon issuance strategies based on various preference tags of interested groups, accurately convert target users and strive to improve ROI.

In the solution of Shence, differentiated coupon issuance strategies are designed based on the various preference tags of interested groups to promote accurate conversion of users and improve ROI through the following two ways.

  • Based on the browsing or consumption preferences of interested groups, targeted delivery of different categories of promotional information

  • At the end of the last hour of the event to send the audience countdown reminder, finally stimulate a wave of transformation

If there is no such tool as Shence intelligent operation, e-commerce operators of brands need to do it manually, which is laborious and inefficient.

Case study: A top consumer electronics brand uses refined crowd operation strategy to establish SMS real-time stimulation program, bringing high value revenue

Compares with attribution analysis found that god policy “contrast goods, watch the live broadcast, the collection”, etc, in the brand “commodity details page to browse” into a single contribution to the highest, then, through the label picture platform, retained analysis found that browse page above 5 times, detailed user retention rates and buy the conversion, a registered product conversion rate is significantly high. Then, during 618, Shenze offered the brand a real-time stimulus package: Users who browse the product details page for more than 5 times in the past 90 days will be sent a direct price reduction message, and those who have not purchased the product within 30 minutes after the key behaviors of “browsing the shopping cart, adding to the shopping cart, collecting the product, comparing the product and consulting the customer service” will be tracked, and the real-time message will be sent again to stimulate conversion.

The final effect was extraordinary. Nearly ten thousand people who met the conditions bought nearly one hundred machines, and the average customer unit price was thousands of yuan. The SMS cost of only a few hundred yuan brought hundreds of thousands of GMV returns.

4. Promote loyal groups: Cultivate brand loyalty through individuation

The brand hopes to use a set of loyalty plans to realize the continuous transition from the buying crowd to the loyal crowd, and make the loyal crowd constantly split to bring more cognitive groups. And the pain point is:

  • How to design a set of indicator system corresponding to the brand’s personalized needs and track the growth of loyal people in real time

  • Bring more cognitive population through the fission of loyal population and evaluate the quality of communicators

Shence can design a personalized index system according to the brand’s unique loyalty plan, and realize the fine operation of loyal crowd. As shown below, badges can be awarded to users who frequently ask questions in the community of a top consumer electronics brand based on the number of times they ask questions. Through the data dashboard of Shence, we can see the users to whom each type of medal is distributed, how many times it is distributed per day, and what kind of follow-up behavior of the users who have received MEDALS.

In addition, god can find the potential KOC through the portrait mining of loyal people, and act as a communicator for fission. The most critical issues of fission are:

  • Who is suitable to be a KOC, i.e. a seed fission user, among the loyal people in the private domain now

  • How much can KOC deliver content? How much of the set transformation can be accomplished by the amount brought in? And then after that, how many secondary transmissions can you make?

Here is one way to measure the effect of fission, where “generation per capita” represents the amount of fission, and “conversion rate with volume” and “fission population value” represent the mass of fission.

Case: An international catering brand evaluated the activity effect and KOC quality through the old belt new fission activity

First of all, in the KOC selection stage, potential KOCs suitable for fission in the loyal population are screened through behaviors and labels, and targeted invitations to old and new groups are sent. Then, in the real-time monitoring stage, the performance of the traffic on the landing page is monitored in real time for the subsequent evaluation of the quantity and quality of fission brought by the loyal crowd. Finally, in the evaluation stage of the activity effect, the achievement effect of the activity on the final goal, namely, the actual conversion number and conversion rate of pulling new and fission, at the same time, the KOC characteristics with high performance were summarized and the KOC selection strategy was optimized.

To sum up, Shence data through three sets of solutions to help brand retail industry private sector business sustained healthy growth, and empower the public domain. This is the end of sharing, thank you for listening.