This sales analysis report is really good.

How to do data analysis projects? How to conduct sales analysis? Following this article, we will learn the methods of sales analysis, improve the ability to analyze specific problems, and solve real-life scenario problems. Recommended for those who are engaged in data analysis.

Many students complain that sales analysis is difficult to do. There is very little data to use, but the expectations of leaders are very high. They always expect to directly improve performance through data. What should we do! Today we will answer it systematically. There are many forms of sales, and it is difficult to discuss without bringing in specific scenarios. So let's look at a specific scenario: an Internet trading platform recruits companies to move in through an offline sales team. At present, the sales department leader has found a data analyst, hoping to do some precise analysis to empower front-line sales and improve sales productivity. Question: How should this data analysis project be done?

1. The key to solving the problem

First of all, what is the key point of this question?

1. Empowerment

2. Improve productivity

3. Accurate analysis

Seeing the topic, many students may be eager to use the "DuPont analysis method". Nine and a half out of ten articles on sales analysis on the Internet will talk about the DuPont analysis method, and some will even add a "disassembly method" at the end to increase the quality of the article (as shown in the figure below).

Indeed, these methods are the basic methods of sales analysis. However, they are not suitable for use here. Note that the salesmen in the title are facing corporate customers, which is a toB business. ToB business means extremely cumbersome follow-up processes and complex customer relationships (as shown below).

This also means that salespeople have to spend a lot of time on traveling, making phone calls, negotiating, attending meetings, drinking with clients, going to bathing centers, etc. In this case, no one will read a bunch of pie charts, line graphs, or bar charts. If you don't believe me, you can count the usage rate of your company's BI system in the sales department. It's great if it exceeds 10%.

Therefore, the real key to this question is: the front line. When no one reads the report, other "empowerment", "assistance" and "precision" are out of the question. This requires us not to think about "what data do I have" or "build an Alpha dog that will make customers place orders as soon as it barks", but to think about "what do sales people really need".

2. The order of solving problems

1. Understand the sales job first

In order to avoid your own fantasies, everyone can understand what these guys are doing through SOPs, interviews, accompanying assignments, etc. (as shown in the figure below).

2. Find the pain points

If you want your data to be "noticed" and the analysis results to be "useful", you must first ensure that what you provide is what others need, and this starts with the pain points (as shown in the figure below).

3. Find the power point

Note that many of the pain points of salespeople cannot be solved by data. But this does not prevent us from using it to attract the attention of salespeople. For example, the most common thing salespeople do is to make phone calls. We can put commonly used customer tags and customer status into the phone book, so that salespeople can see at a glance which are old customers they haven't contacted for a long time and which are new customers who have just signed contracts and need to be followed up. It is convenient to make phone calls, and the report usage rate will naturally increase.

Note that you need to control the intensity of your efforts here. Take the phone book as an example. Some product managers have also paid attention to this point, but they particularly like to add a "work plan" before making a phone call and force the front-line staff to fill out a work plan. They also call it: You see, the first step of the sales process is to list the customer list, so you need to make a work plan, which is scientific and reasonable.

As a result, the product usage rate naturally dropped to 0! Because only the product manager who writes in front of the computer every day likes to make Gantt charts and write daily plans. If you ask him to go to meet customers in the hot sun, wearing a tie, sitting in a Didi, and being stuck in traffic for several hours, he will not have the heart to write such flowery things. Only a fool would write such things. For all things facing the front line, the requirements for convenience are far greater than scientificity. Remember this.

4. Look for opportunities for improvement

When our data is read by others and our data products are used by others, we can think about how to improve them. Note that when facing the front line, remember not to talk about high standards. This is simply asking for a fight, and it is easy to get a response like "If you can do it, do it" (as shown in the following picture).

If you find a problem, it is best to tell the front line directly: this works! Using the example above, if you really find this problem, you can first look at what people with a high signing rate have done. If we find that showing cases of peer customers is effective (improving by about 5%), then we can directly build the industry case library into the product as a feature. Tell them: When chatting, click this button, it works!

The effect is definitely much better than pie charts, line graphs, and bar charts. Sales are also willing to listen and use it.

5. Gradually promote its use

It is possible that we have discovered 100 problems through data, but if we want to solve them, we have to solve them one by one. It is best to push the next one after seeing the improvement of one data. Many data product managers like to come up with a very complex "comprehensive solution" for the front line at one time, but in the end, it naturally lies in the mobile app/applet and gathers dust.

Some students may say: I don’t have the opportunity to make data products, what should I do? The biggest difference here is not whether you can make a data product, but whether you have a real business problem in mind, or a pie chart, line chart, bar chart, addition, subtraction, multiplication, division and square root. If you have a real problem in mind, we can:

1. When making a PPT, directly make the "Sales Winning Strategy", write the data findings and corresponding solutions together, and package them for output.

2. When doing Excel, mark the problem directly on the change curve, and write a note next to the personal highlight that can overcome the problem in the detailed table.

3. When making an oral report, after talking about the data that reflects the problem, directly tell a personal story about how the problem was solved and tell everyone that this can improve the data.

Even if you don’t have a product, you can still say something that others are happy to hear. You can even promote the company’s project and get yourself an opportunity to output data products.

If you only have pie charts, line charts and bar graphs in your mind, even if you really develop a data product, the final result will be: the report opening rate is 5%, and only the sales team leader and the sales data statistician will read it. After reading it, they will say: I knew it a long time ago, what's the use of this...

3. This scene can continue

for example:

1. The company background is changed to: toC, traditional beauty salon

2. Company background changed to: toC, automobile

3. Offline team changed to: Telephone sales

4. First-line salesperson changed to: Regional manager

5. First-line sales to: sales operations

6. The product sold is changed to: advertising service

7. The products sold are changed to: data products

You can think about what difference these seemingly small changes will make and what the data analysis project needs to be like to meet the needs.

Some students may ask, what is the significance of this kind of scenario deduction? Because in enterprises, real data analysis has to face all kinds of strange scenarios.

for example:

The service objects may be: front-line employees, middle-level leaders, senior leaders

Business content may be affected by the epidemic. A large number of traditional beauty salons, gyms, and auto 4S stores that have no transactions with the Internet have also come to engage in online traffic diversion and online live broadcasts.

After the Internet platform itself promoted the entry of enterprises, it then promoted data products, marketing services, office OA, and data middle-end resources...

All of these are real events that happen around us every day, and they are real problems that need to be solved.

Without getting into specific scenarios, when encountering problems, you will just ask everywhere "How to analyze Internet data???" and the answer you finally get is still the unchanging AARRR, DuPont analysis method, or funnel model - it's useless.

Therefore, if you really want to make data analysis useful, you must carefully consider specific scenarios and improve your ability to analyze specific problems.

Author: Down-to-earth Teacher Chen

Source: WeChat public account: "Down-to-earth Teacher Chen (ID: gh_abf29df6ada8)"