Super detailed real-life examples of user portraits

In today's data-driven business environment, user portraits have become an important tool for companies to understand customers and optimize products and services. However, how to effectively use user portraits to diagnose and solve practical problems remains a challenge. This article uses a real case study of a high turnover rate in the sales team of an Internet company to explore the correct application of user portraits.

User portraits are something that many people talk about, but every time they encounter a problem, it seems that they stop there except for listing a percentage of "gender, age", making it difficult to truly diagnose the problem and take action.

Today, let’s take a look at a specific example of how to use user portraits to diagnose problems. Without further ado, let’s get straight to the point.

Problem scenario: An Internet company uses traditional telephone sales to promote its advertising business. Recently, the three-month turnover rate of new sales employees has been extremely high, so high that the leadership cannot stand it anymore. They ask to see the user portraits of the sales staff and analyze where the problem comes from.

Q: How should I read this portrait?

1. Common Mistakes

How many students, seeing the words "user portrait" in the title, start to dig up sales data, such as gender, age, education, years of work experience, number of follow-up leads, number of transactions, performance... a bunch of indicators are posted on the PPT.

Because sales are internal employees, they enter their resumes when they look for jobs, and there are many fields in the resume that can be used. So this "user portrait" PPT is likely to be very long...

The question is: does this explain the problem?

Obviously not!

2. Ideas to break the impasse

The problem is obvious: this is just listing data and does not address the issue of "sales resignation" itself.

Regardless of the data, why do sales people leave their jobs? There may be many reasons:

● The company is just rubbish and has a very bad reputation in the industry

● HR recruited the wrong people from the beginning, just filling the gap

● Poor sales management and poor employee experience

● The sales performance is not good and the sales cannot make money

● You can make money from sales, but it’s too hard work

Moreover, these factors may overlap with each other. For example, because it is difficult to sell advertisements, management tends to use the human wave tactic, forcing employees to work overtime; forcing employees to work overtime reduces the work experience and increases the psychological burden of employees, forming a vicious cycle.

Therefore, from the perspective of analyzing the problem clearly, you cannot "find nails with a hammer", because the leader said user portrait, I have to user portrait everything. First sort out the analysis logic, and then see what problems the user portrait can solve.

3. Constructing the analysis logic

First of all, the people who leave the company are likely to say, "This company sucks!" So reasons like "The company sucks" are the easiest to rule out first. A truly sucky company, judging by the data, is really sucky from beginning to end (as shown in the figure below).

Secondly, poor management is also relatively easy to rule out. As long as the company as a whole is not bad, then some sales teams must be poorly managed and some are very good. If you look at the turnover rate from the team's perspective, you can easily find the problem (as shown in the figure below).

Again, for example, during certain time periods, the quality of sales leads is too poor, which interferes with sales work. This can also be eliminated in advance (as shown in the figure below).

By stripping away the obvious macro issues, we can focus on the ability of sales itself. Detailed discussion: Is it that the business itself is difficult to do and sales can't handle it; or is it that sales can actually be handled, but HR hired the wrong person and the person they hired is not capable enough. The judgment here involves the evaluation of sales ability and the basic characteristics of personnel, and user portraits can be used to assist.

4. Entry point of user portrait

User portraits are not just basic features such as "gender and age". Various user behaviors and social relationships are also important components. When conducting specific analysis, we must pick out the features that "can best explain the problem". As for the resignation issue, the most powerful factor is income - who would not leave if there is no money to be made?

Note: The problem scenario is "new employee turnover rate within three months", and general income is settled on a monthly basis, so it is necessary to distinguish the three-month situation. It should be noted that it is best not to just calculate the average here, because it is very likely that some people will leave without opening an order in the first month (confidence is hit), and some people have a high month and a low month (income is unstable and they have no confidence to continue working), so it is best to divide it more finely (as shown in the figure below).

With this step, we can distinguish between high-income and low-income people, and combine it with the turnover rate to explain the problem to a large extent. It is very likely that most of the people who leave are low-income. Moreover, if a person leaves a high-income job, it is very likely that they find it difficult to do the job and can earn more by doing the same amount of work in another company, which can be further analyzed.

5. Discuss from the basics to the depths

If we want to go deeper, we need to further explain: what are the reasons for low income?

There may be the following reasons:

● Not enough leads are assigned (management issue)

● Enough clues were assigned, but not completed (execution problem)

● Enough leads were allocated, but they were completed but not converted (skill problem)

● There are enough leads assigned, they are completed, there are conversions, but the orders are too small (still a management problem)

Here we can simply use the matrix analysis method to distinguish the situations. Generally speaking, if there are problems with the lead allocation mechanism itself, such as uneven allocation, too much concentration of good leads on certain people, and no differentiation by industry, which leads to some people's orders being small... these cannot be classified as sales problems, and can be improved by optimizing the internal management mechanism (as shown in the figure below).

In this way, the discussion of whether HR did not recruit the right people will be very focused. By comparing the characteristics of those who have no execution ability and lack skills with those who have execution ability and are willing to do the work, we can find out what the real sales characteristics are. And then reversely infer whether the quality of the people recruited by HR is not high.

At this point, you can take out the user characteristics and compare them. In theory, age, gender, region (whether it is a dialect area), education, years of experience, these fields that can be directly obtained from the resume can all be used. At the same time, you can extract "whether you have Internet advertising sales experience" from the text as a label for analysis. It is very likely that HR recruited all the people with the word "sales" in their names, but many people could not adapt to this sales model and thus left.

6. A special case

There is a very special situation, that is, after doing a lot of comparisons, I found that:

*There is no big difference in the data!

*There is no difference in income!

* No difference in execution!

* There is no difference in features!

Some people quit their jobs without knowing the reason, but there are many who don't make any money even though they haven't quit!

What to do!

If this is true, then it is very likely that this business has a clear newspaper boy effect. That is, how many newspapers a newspaper boy sells has nothing to do with his personal characteristics, sales skills, or user needs. No matter how hard he tries, only two out of 100 people will buy.

If this is the case, then there is another strategy. Sales will be lost anyway, and there is no way to improve it, so cut recruitment costs and find ways to get more cheap labor.

VII. Summary

The above is just a brief demonstration of the analysis ideas. The actual work may be more complicated, but it adheres to a basic idea: doing data analysis is not like using a hammer to find nails, but it is about looking at the food and analyzing specific problems specifically, with the goal of solving the problem.

Each tool must be organically combined with the actual situation. User portrait is a combination of a set of labels, and it is necessary to choose the appropriate label according to the situation, or even develop new labels.