How does a data analyst write a "useful" analysis report?

Many data analysts are trapped in the endless collection of data, but these data may not be used, or even have no effect. Faced with this situation, how should we deal with it? The author summarizes the ideas of how to write a "useful" analysis report, hoping to help you.

Preface

Many data analysts will be caught up in the endless data collection needs of the business side, which may be temporary data collection that has no use and is just for review; there will also be endless reporting needs, but the output reports are likely to be discarded after being viewed once.

At this time, the data analyst will fall into self-doubt. How can I, a data analyst, do these repetitive and "worthless" data collection work every day? I am just a data collection tool for the business side. And your leaders and the business side will feel that you have not created any business value and that your work is not done well, even if you have been working overtime every day to meet the endless needs of the business side.

So how should you break the deadlock at this time? You can try to write a data analysis report on the business, point out some problems in the business and give optimization suggestions, and do everything possible to implement your suggestions so that they can really improve the business.

Then the business side will understand your ability, and you can negotiate with them, asking them whether they want you to spend time on meaningless data collection or conduct valuable data analysis. As long as the business side is not a fool, they will definitely make the right judgment. In this way, you will gradually have the right to speak and be able to do more of what you want to do. So how to write a valuable data analysis report? Let's see the following introduction.

01 Analysis report function

A good data analysis report can answer the questions that the audience of the report encounters. Let's break down the purpose. The audience includes ordinary business students, business line leaders, and even company management. These people encounter various problems in their daily work. You need to select valuable questions from these problems to analyze the reasons behind them.

Generally speaking, questions from ordinary business students can be ignored directly. Although this may sound a bit discriminatory, because their perspectives and information inputs are limited, they rarely find valuable problems worth analyzing. As a data analyst, you generally have to serve business parties who are several levels higher than you.

One more thing to add, why is the data analysis report not about discovering the good points in the business and making them flourish? Because the good points of the business side are not yours to discover, they have already reported them to various places and tried their best to land them in other applicable places. Only when they can't solve the problems themselves can you give them some input and guide them on how to make breakthroughs, otherwise how can they let you teach them how to do things. This is human nature, and you will get used to it slowly after spending more time in the workplace.

02 Topic Selection

There are all kinds of problems in daily business, and not every problem is worth in-depth analysis. This requires you to have a strong business sense. Of course, this judgment will be difficult, and you need to understand the industry and the company's business operation logic.

Here is a simple and rough method of judgment, that is, the higher the job level, the more problems that bother people are worthy of in-depth analysis, because their energy is limited, they will only think about the business modules they care about, thus causing problems. Generally, the higher the job level, the more information they have and the stronger their ability, which is equivalent to naturally helping you to make a relatively accurate filter.

Another point is to see if there are theoretically solvable resources and solutions to this problem. For example, if you cannot solve the problem caused by policy changes, you can only passively accept it instead of trying to analyze something and then solve the problem. You need to find a problem that your resources can solve and analyze it. After selecting a meaningful proposition, you can use all your skills to analyze it.

03 Conclusion

The most important thing about a data analysis report is that it must have a conclusion and express your point of view. However, the content such as XXX index increased by 10% month-on-month is not a conclusion, it is just a statement of fact. You must give the corresponding data indicator results that best reflect your analysis proposition, what is the business meaning behind the data results of this indicator, and give your feasible suggestions.

Of course, this is very difficult to accomplish, and it is often a big hurdle in determining whether a data analyst can grow into a senior analyst who drives business growth.

This reflects the comprehensive ability of a data analyst. You must understand data as well as operations, products, marketing, research and other business positions. Only then can you find out what the business problems behind the data are and give reasonable optimization plans based on the actual situation of the company.

Only by giving such a conclusion can it be a complete data analysis report, which can bring real help to the business and achieve the first step of data-driven growth. You need to understand the difference between you and the business side, and what you can do that they can't do, in order to reflect your value.

04 Content

4.1 Typography

It is strongly recommended that you write your data analysis report in the form of general-specific-general. Although this is the opposite of your analysis process, many analysts just write down their analysis process and think it is a data analysis report. In fact, this kind of layout is very bad.

You must use the general-specific-general format. The beginning of the report should be a very attractive conclusion that will attract people to read on. Your audience are all relatively high-level business parties, especially management, who have very limited time. If you cannot arouse their interest at the beginning of the report, they will most likely skip the report directly. This will be very disadvantageous to you. At the very least, you will not be able to obtain the corresponding resources. At worst, their impression of you will be worsened, and they will think you are doing some meaningless research.

Then the total score to total is very consistent with the reading habits of most people. I believe that when you were in school, your teachers would also ask you to write essays in the form of total score to total, which would make it more likely to get high scores. The essence behind this is the same.

4.2 Data display format

If you can use charts instead of tables to display data, you should choose appropriate charts, and don’t use too many colors in the charts, which will affect the visual perception. Here I suggest that you learn the chart design language on the AntV platform. There are many systematic summaries about the design of charts. This is the accumulation of many years of Ant Group, and there are still many references. It is recommended to deeply experience the design aesthetics in it.

In addition, if summary indicators can be used, detailed indicators are not needed. It is not advisable to display too much detailed data in the report. The length of the report is limited, and readers do not need to fully understand your calculation process. You need to establish your data expertise elsewhere to make them trust your data.

Data analysis requires a full chain, so not only the result indicators should be displayed, but also the core process indicators.

4.3 Logic

Your analysis logic must be rigorous and closed-loop. The arguments in the data analysis report are clear, and it is better to be concise rather than verbose. Your arguments are clear and reasonable, and then you draw your conclusions through rigorous logical reasoning, rather than making a lot of self-righteous guesses. Only in this way can the conclusions of the analysis report stand up to scrutiny and everyone will recognize your conclusions.

05 Landing

Many data analysts fall into a misunderstanding, thinking that their work is over once they have written the analysis report, and that whatever happens next is none of their business.

A data analyst is qualified only if he can promote the implementation of your business suggestions. It is very difficult to implement a project, but it is precisely in this way that your value can be reflected and it is also the core competitiveness of a data analyst.

Technology is changing with each passing day, and you will never be able to compete with your latecomers in this regard, because the learning threshold here is extremely low. Only this basic ability is the moat you need to build.

How to implement your data analysis report? First, you need to build trust in your work ability in daily work, such as accurate data collection, easy-to-use report design, etc., to establish the first wave of goodwill. In addition, you need to prove the importance of your analysis report to your leaders and business leaders. In the end, the adult world is all about interests, that is, what benefits can they get from implementing this analysis report? First, convince your leader to approve it, and ask him to help you convince the business side to improve the core business indicators of the business side. After reaching an agreement with the business leaders, things will become easier.

Next, you need to implement your suggestions together with the business party you are cooperating with. Whether the business party will work together with you depends on their interests. What's the benefit for them in doing this? The leaders eat meat, and they always have some soup to drink. You need to give some credit to the other party appropriately, so that a project can be truly implemented. Otherwise, even if the other party's leader agrees to this period, it will still be difficult if the other party gives you some obstacles in the process.

Then after all your hard work, you have put the analysis report into practice. Then the real value is generated, and data-driven growth is achieved in the true sense.

06 Last

The data analysis report is the best performance of a person's comprehensive data analysis ability. You must reject more invalid demands from the business side and spend more time on meaningful things, so that your sense of achievement will be higher. The business side will also trust and rely on you more and more. Satisfying all the needs of the business side is never the result they really want, and they will not remember your kindness. Instead, you will gradually become a "cousin" or "cousin sister", thus falling into a vicious circle.

I hope that everyone can gain something after reading this article. I also hope that everyone can go further and further on the road of data analysis and maximize their own value!

Author: A Kun, member of the "Data Creator Alliance".