How to create a B end customer image?

1 .List the underlying data.

Before collecting data, according to the purpose of using the enterprise user portrait, the basic data required to construct the enterprise user portrait are listed from the perspectives of macro, meso, and micro layers.

The macro layer refers to the industry attributes of enterprise users, and different industries have different market structures, operation modes, and operation rules. Through the characteristics of the industry, understand the current situation and development trend of the target enterprise user industry;

The meso-level refers to the enterprise attributes of enterprise users, such as the establishment time of the enterprise, the scale of the enterprise, the scale of personnel, the scale of income, the scale of active users, and the evaluation of use. Understand the current situation of target enterprise users through enterprise characteristics;

At the micro level, the user portrait of the enterprise is not only the characteristics of the enterprise attributes, but also the characteristics of the role attributes within the enterprise, that is, the role characteristics of the decision-making chain, which can be divided into: decision-makers (bosses, executives) and users (employees), and the concerns and needs of decision-makers and users are very different.

2.Survey Subject.

After listing the basic data, it is necessary to determine the object of the research (collection) data, and only the role in the decision-making chain has the research value.

For example, the existing B-end product of "enterprise welfare system" is to assist enterprises to establish a positive incentive system and solve the problems of high cost of employee welfare procurement, difficulty in distribution, and low employee satisfaction.

3.Research methodology.

Once you've identified the research object, it's time to investigate. Enterprise user research can use a combination of qualitative and quantitative analysis.

Qualitative analysis refers to the method of analysis by researchers through interviews, insights, etc., and based on industry research reports, past experience, etc., to mine the behavioral motivations, needs, changes and other data of enterprise users;

Quantitative analysis refers to the research method based on data and possibilities, comparing and analyzing various indicators, characteristics, and interrelationships of enterprise users, and using data to test certain hypotheses.

If you need to create a B-end user portrait, you need to do the following 4 things:

1. Clarify the target users (enterprise information).

2. Clarify the decision-making chain (decision-making chain role, right to speak, basic characteristics) 3. Sort out the user path.

4. Refine commonalities (scene stories).

There is a big difference, and I will explain the difference between a B-side product manager and a C-side product manager from these three points.

First, the value perspective.

Second, the market perspective.

3. Design perspective.

Hope it helps.

The difference between the customers on the other side and the customers on the C side is that the disadvantages are mainly for some enterprise customers who are similar to companies, uh, and then the customers on the C side are individuals.

At present, there are many third-party car appraisal platforms on the market, but most of them are mainly C-end and provide appraisal services for individual buyers. The car e-valuation mainly serves B-end customers, such as car rental and auto finance customers, who need not only an independent and authoritative appraisal report, but also improve the efficiency of business processing.

Home page collation of the task characteristics of the boss of the human resources business

This is very important, the boss, the human resources industry, and a series of other characteristics, through conversation contact to understand.

Product Operations

Business owners, especially human resources, have a clear positioning, if they go high-end, the packaging of the product is very important, UI, to the packaging.

If the low-end needs to have characteristics, because human resources, HR bosses are the best at finding characteristics, and it is often difficult for them to enter their eyes.

Market Operations

This is to do national or local according to their own situation, normally the market operation according to their own products, need to be planned and launched, to test, according to the test and conversion of continuous adjustment, online or offline.

Brand operation

This aspect of brand operation is very large, generally speaking, we will look at two situations, one is the product sticker brand, and the other is the brand sticker product. The former pushes the brand plan according to the product attributes or related characteristics, and the latter is to make a virtual IP similar to the corporate culture, and any product can be pasted.

To look at the situation specifically, two simple sentences are too big to ask. Every question is a three-no-book.

The B-end generally refers to the enterprise merchant side, which performs good products and reflects good services

Although the operation of the C-end and the B-end is very different, and many of the C-end operation routines will not work when they reach the B-side, but the side Hongyuan is used by individuals in the end, and the products and users that need to be paid attention to are still the products. Specifically, it is a product that is absolutely stupid, and it needs to be analyzed in detail.

01 Goals.

Like to C operation, before you start to implement the operation, you need to find the core goal of your product, that is, the North Star indicator.

Without a clear goal, effective implementation in Chanchang is essentially increasing costs.

02 The whole process of conversion.

03 Business model.

04 Platform-based business.

What should be done for platform-based business operations?

c has a definition of: consumer, copy client; If you take "consumer", it means a consumer attacker, an individual user, or an end user, and the client is used. For example: NetEase News, NetEase Cloud** and so on.

b Definition is: business; System-based software, tools, or platforms that are typically used by businesses or businesses for work or business purposes. For example: NetEase Cloud, NetEase Youshu or internal ERP system, etc.

The essence and characteristics of C-end & B-end products:

In the Internet industry, the end and the C end are actually abbreviations.

B is the acronym of the English word "business", and the B side refers to the enterprise user business, as the name suggests, it is the service product provided for merchants, enterprises, and business departments, and it is an indirect service to users.

C is the acronym of the English word "consumer", and the C side refers to the individual user consumer. As the name suggests, it is a product that provides services to individual users and directly serves users.

Like we often say B2B, B2C, C2C, according to the merchant-user is easier to understand!

Hello, C stands for the customer and B stands for the merchant.

Therefore, B2B is often referred to as business-to-business, such as wholesalers on Alibaba, facing manufacturer users;

The often said B2C, merchants to users, such as JD.com and Tmall;

There's also C2C, where individual sellers sell to individual sellers, for example.

In the Internet, the C-side generally represents the buyer or consumer, and the B-side generally represents the seller or merchant.

The C-end should refer to the end consumer group; The B-end is the channel customer.

Enterprises need accurate user data to build user portraits, and the data collection data is roughly divided into four categories: network behavior data, in-service behavior data, user content preference data, and user transaction data. Network behavior data: in-service behavior data such as active people, page views, visit duration, activation rate, external touchpoints, social data, etc

Browsing path, page dwell time, depth of visit, unique page views and other user content data: browsing Favorite content, comment content, interactive content, lifestyle preferences, brand preferences and other user transaction data (transaction services): contribution rate, customer unit value, joint rate, return rate, churn rate, etc. Of course, the collected data will not be 100% accurate, and they are all uncertain, which needs to be modeled in the later stages to judge, such as a user who fills in the gender column. However, there is an 80% probability that her gender will be "female" based on her behavioral preferences.

It is also worth mentioning that it is best to store user behavior data at the same time as the scenario in which the behavior occurred, so as to better analyze the data. Behavior modelingThis stage is to process the data collected in the previous stage, and carry out behavior modeling to abstract the user's labels, this stage should focus on high probability events, and exclude the user's accidental behavior as much as possible through the mathematical algorithm model. This is where machine learning is also used to guess the user's behavior and preferences, like an algorithm of y kx b, where x represents known information and y is the user's preference, which is accurate by constantly exactizing k and b.

At this stage, a number of models are needed to label users. The user car model judges whether the user has a car and whether he is ready to buy a car according to the user's attention to the topic of "car" or the purchase of related products, and the user loyalty model judges the user's loyalty, height, body shape model according to the user's purchase of clothing, shoes, hats and other supplies, and judges the literary youth model according to the user's speech, comment and other behaviors, and judges whether the user is a literary and artistic young user, and the value model judges the user's value for **, which is very useful for improving user retention rate (e-commerce** generally uses RFM There are also many models such as spending power, default probability, churn probability, and so on. Therefore, to sum up, data quality + user scenario + model establishment = accurate user portrait, there are many professional data platforms for user portraits in China, such as the intelligent customer data platform of Chuangluo Technology, which can build a user portrait foundation for almost all data types of enterprises, including multi-channel data integration, real-time data collection, and customer data tag sets based on enterprise marketing scenarios to achieve accurate portraits.

There is a lot of work that needs to be done before defining user tags, first of all, you need to determine the data**, and do a good job of metadata and master data management, and then collect, organize, clean and transform the data according to business needs to ensure the quality of the data.

Use Photoshop's lasso tool and then blur it, it's too fine, and no matter how good it is, there are experts who can see it.