I prefer to play video games. I am a vegetarian and fun-loving type.

In order to improve their skills, they will watch some videos of masters. Every time after watching the master video, I have a surge of blood, quickly ready to go on the number of points.

However, reality tells me that watching others play 6, their own hand is still not good.

It’s not just the hand speed of the operation, but the fact that you can’t copy the situation in the instructional video and don’t know how to deal with different situations.

He will play a whole game, and then often talk about his own thinking and judgment of the current situation, and then give his own solutions.

After watching this kind of video, I usually improve my level just a little bit.

I found this experience very similar to when I was learning to write analysis reports.

In order to learn how to write reports, I read many excellent reports from third-party data platforms. I also read a lot of dry articles to understand the structure of the report.

But when they really write, can only restore its form, it is difficult to restore its god. The final report looks like nothing but the framework and the content doesn’t stand up to scrutiny.

Later, I came to understand that the core of the report was the logic of the content, and the writing of the final report was as simple as filling in blanks.

So, how to write a report is not to tell you what the style of the report is, but to tell you how to comb through the logic of the report.

I’m going to talk about that today.

Order of thought for writing a report

When writing a report, many students will not think about how to write, but directly write down the process of their own analysis. Because of themselves, they feel that this is the way they analyze problems, and it’s natural to re-present this way of thinking.

But we need to understand what the purpose of the report is.

A report is not a paper, and you don’t need to step by step to give your colleagues a sense of the rigor of the conclusions. The logic problem is the problem you need to solve as a data analyst.

In essence, report is a way of internal communication. Communication in an enterprise is about efficiency. Report should let the other party know what you want to say quickly.

To analyze a problem is to break it down step by step and come to a conclusion.

In order to communicate the information effectively, you should first tell the other person what your conclusion is, and then tell the other person why you came to that conclusion.

So, the steps of writing the report and the steps of analyzing the problem are reversed.

If you reverse this thinking, you can think about the logic of the report and it will be much clearer. This is something that no one can do, but no one can do.

The first step is to clarify the purpose of the report

First, think about the purpose of the report: What is your report really about?

The criteria is, what are people going to do after they read your report? Or what will change?

If people read your report, and there’s no follow-up or change, and people keep dancing and playing music, there’s no reason for the report to exist.

The value of a report is measured by whether there is follow-up to the report and the incremental business resulting from that follow-up.

Data analysis is part of the decision-making process throughout the enterprise. Decisions go through three stages: what, why, and how, so a data report has nothing more to say than these three things.

The first is “what?”

This kind of report is to tell the business person what the problem is, and the follow-up action is to analyze the “why”

Such reports are usually regular periodic reports, such as daily, weekly and monthly reports. Such reports often list changes in operating indicators and show what is wrong with the business.

Because business development encounters a variety of problems every day and analysis resources are limited, the purpose of this type of report is to let the business understand the current situation and provide feedback on which problems need to be addressed urgently. The analyst can prioritize those issues of concern to the business. This is the follow-up to the report.

For example, the decline in sales may be caused by four factors: ABCD. It may take a lot of time to analyze all ABCD clearly, so at this time, the four factors of ABCD can be made into a report and synchronized to the business side. The purpose of this report is for the business side to output some business understanding to determine which subproblem in ABCD will be solved first.

The second “why?”

This type of report tells the business people what the cause of a business problem is, followed by an analysis of “what to do”, or the business people directly conduct various business attempts.

This type of report is common because finding the cause is the specialty of data analysis, but finding the solution is not the specialty of business people.

If the data analysis finds that the decrease of turnover is due to the decrease of conversion rate, then the decrease of conversion rate is due to the promotion activities of competitive products in a certain region, which leads to the decrease of local conversion rate. Such reasons are enough to make business people think about how to deal with them, and data analysis does not have a high say in this aspect.

The third “what to do”.

Such a report tells the operational staff exactly what to do and is followed up by the implementation of the programme.

This part is easy if you have developed very mature recommendations from the previous analysis. The analytical approach to the problem can be found in previous articles in the Analytical Thinking series.

The second step is to support your conclusion

When we analyze a problem, it is unlikely to find the root of the problem directly. We often have to take some detours, eliminate some wrong answers, and finally find the cause of the problem.

But along the way, we find some unexpected information.

So when we write the report, we will want to include these unexpected information in the report, otherwise it will feel too pity.

But this can lead to a report that is out of focus, and the reader gets distracted by the scattered points of view and loses track of what is being said.

Some are even more egregious. I’ve seen some rookies write analysis reports. A PPT of 40 to 50 pages is very comprehensive, and all the conclusions drawn in the analysis process are presented. For example, when faced with A problem, three hypotheses of ABC were proposed, and then A was assumed to be wrong, and then the hypothesis was verified and B was also found to be wrong. Finally, hypothesis C was proved to be true, and then three hypotheses of EFG were proposed because C was true, and the previous process was repeated. The readability of such a report is very poor, maybe twenty or thirty pages. No one knows what you’re going to say.

Therefore, the second step is to find out what arguments we can support the big conclusion after determining the core conclusion required by the report objective. We can ignore small conclusions that are not explicitly related to the core conclusion.

Don’t feel bad, some small points may be wonderful, but your presentation is wonderful not because some points are wonderful, but because they are combined in harmony.

As to which arguments you need to prove the conclusion and whether the arguments are sufficient, you need to have a strong systematic, structured and logical thinking, which is welcome to read my previous related articles.

• Logical thinking
• Structured thinking
• Systematic thinking

Step 3: Data to support your argument

The third step is actually a continuation of the elaboration of the previous part.

The last step is to figure out what the arguments are that support the conclusion, and the next step is to figure out what the data is that supports the argument. The principle is similar to the previous step, so I won’t repeat it here.

Let me just say a few things about data selection.

Data analysts tend to look at detailed data when analyzing problems. Because detailed data has more information, more problem points can be identified. For example, if we look at the sales of two different products, are there differences? The end result might be something like this.

This graph shows that category A generally has higher sales than category B. But this graph has redundant information. As can be seen from the figure, the sales of the two categories are stable, so the time dimension actually plays no role in comparing the difference of the transaction value of the categories. So we don’t have to include the time period.

Once the time dimension is included, readers of the report are left wondering why sales rose one day and fell another. This has nothing to do with our argument that category A generally sells more than category B.

So this data we can hide the date dimension, using the average, using two bar graphs.

At this point, the main logical part of our report is complete, but it is not enough.

Step 4 how to make the business take this report seriously

The point of your report is now clear to the manager. But no matter how logical the advice is, it won’t matter if no one is interested in it.

So we have to think about the question, why should anyone else take this report seriously?

Put this step last because only if you know what the report says will you know who the conclusions are useful to and what their relationship is?

After thinking about it carefully, there are several ways to increase interest in your presentation.

The first is the title of the report.

If your report is about how to increase sales of a product, the title of the analysis might not be as appealing as “Data Analysis Report on How to Increase Sales of a Product”. Instead, it might be “Three Suggestions for Increasing Sales of a product”. When students read the report, they will know exactly what you are trying to say, and they will know that the report is not about generalities, but has specific suggestions that are worth reading.

The second approach is to introduce the background and severity of the problem at the beginning of the report.

You can use the SCQA model from pyramid Principles.

The third way is to explain the benefits of adopting the proposal.

Say you tell them you can expect a 10% increase in sales if you follow this advice. The business people will be more aware of the actual value of the report and more willing to listen to your suggestions.

conclusion

Reading other people’s excellent reports will tell you at most how a good report makes one thing clear. But looking at other people’s reports doesn’t tell you how to sort things out first.

After thinking in accordance with the above steps, the framework part of the report idea is sorted out. With this narrative framework, learning about presentation formatting, graphic visualizations, and so on, will get you twice the result with half the effort. In fact, more than 70% of your report writing is organized.