It’s 2024, here are some data-driven methods

In business competition, data has become the key to business operations. This article will share some practical methods to achieve business growth driven by modifying the way data is expressed, such as changing the wording, adding attributives, changing logic, and changing indicators. However, it should be noted that true data-driven development needs to be combined with business reality and based on real and reliable data. Let's take a look at this article.

Friends who often do data analysis know that there is only one fastest way to improve performance, and that is to change the caliber.

Changing the caliber can be completed in a moment, and only one data analyst is needed, so the per capita output is extremely high .

Compared with other methods, which require a lot of manpower and material resources and take half a year to be effective, this method is much more efficient.

If business colleagues want to improve performance growth, they also need to rely on different strategies such as product-driven, marketing-driven, etc.

But data analysts are very simple and direct. They just modify the data and use "data-driven" methods.

Here are some ways to use data-driven methods.

1. Adding an attributive

The first “data-driven” approach is to add an attributive.

Anyone can be the first as long as more adjectives are added.

In the past two years, the new energy market has released new cars at such a fast speed that it is hard to keep up. In order to make people remember their cars, manufacturers have to keep them in the public eye, so various rankings have emerged.

For example, Ideal released a new power brand sales list , and Ideal was far ahead on the list. Why? Because this list is limited to "new power". Since Ideal was far ahead on this list, it has released this list almost every week, which has rapidly increased Ideal's brand awareness.

However, Ideal's ranking on the list is not particularly outrageous . After all, the new forces still have many competitors, and getting first place among the new forces is indeed a relatively good result.

The following one is a little bit outrageous.

A certain picture released a poster after the new car was launched in 2021, and took the first place in a certain market .

The adjective "first" is long enough. Let me read it to you: "With an average brand guide price of over 320,000, it ranks first among mature car companies and Chinese brands." It has 24 words in total, including 320,000 average brand guide price, mature car company, and Chinese brand . There are at least three adjectives. Under these limited conditions, there are basically no competitors left.

Last year, joint venture cars had a hard time. Toyota and Volkswagen, which used to be the top sellers, both saw a sharp decline. As a result, Toyota had to start adding adjectives to its posters. Last February’s sales figures only had the new title “joint venture” written on them.

Such examples abound. In addition to the automotive industry, the mobile phone industry is also full of lists with such qualifiers.

In short, as long as it is not the first, add an attributive.

As long as there are enough adjectives and few enough competitors, you can always find the first place.

2. Change indicator logic

The second method is to modify the logic of the indicator.

This practice is not only done by data analysts; it is also quite common in financial analysis, which is known for its rigor.

For financial analysis, since financial indicators are three fixed tables, the indicators themselves basically cannot be modified, but the calculation logic of the indicators can be modified.

For example, in 2013, Anshan Iron and Steel Group was under great pressure to make profits. The company had suffered losses for two consecutive years, and if it suffered losses again that year, it might face the risk of suspension of trading or even delisting.

On April 9, 2014, Anshan Iron and Steel released its financial report, announcing that it had achieved positive profits in 2013, a year-on-year increase of 119.13%, successfully avoiding losses.

So, how did Anshan Iron and Steel achieve this turnaround?

In fact, the answer is "data-driven".

In those years, the steel market was in a long-term downturn, the market environment was very poor, and it was generally believed that it was almost hopeless to turn losses into profits. Anshan Iron and Steel was able to turn losses into profits by adjusting the depreciation period .

Anshan Iron and Steel extended the depreciation period of houses and buildings from 30 years to 40 years , and the depreciation period of mechanical equipment and transmission equipment from 15 years to 19 years. In this way, the annual depreciation cost was diluted.

After the adjustment, Ansteel's net profit in 2013 increased by about 900 million yuan compared with 2012, reaching 770 million yuan. In other words, if the depreciation period is not adjusted, Ansteel's net profit in 2013 is actually still a loss.

This practice is not limited to one company. Many companies, including Sangang Minguang, Shandong Steel, Fuchun Environmental Protection, Fangtai Special Steel, and Hebei Steel, have used this method.

The business itself has not changed, but the caliber has been changed, and the business has turned from loss to profit. Tell me if data-driven is powerful or not.

3. Change indicators

The third method is to change an indicator.

For example, a new product feature was launched with the original intention of increasing the user's usage time. However, after the launch, it was found that this feature did not affect the user's stay time as expected.

What should I do at this time?

So let’s not look at the length of stay, but look at other growth indicators. After a lot of work by the data analyst, we finally found out from more than 100 indicators that after the product was launched, the number of active users among Android users opening the APP increased. So the launch of the product function was finally a success.

The core of this approach is to find a satisfactory indicator. If a satisfactory indicator cannot be found, it means that the data analysis is not deep enough and the hidden information has not been found.

Some situations are more direct. Some leaders assign tasks such as "analyze the value of this function", which means that the leader needs to find a positive indicator to report, and if it is not a positive indicator, the analysis will continue until it is positive.

I even had a friend who told me that their leader directly asked them to find an indicator that could show a 30% growth, and then asked the data team to collect such data. I was really shocked when I heard that.

These are some common "unspoken rules" in the field of data analysis.

4. Fake Data

If you think the previous "data-driven" method is too troublesome and you have to find indicators and change the caliber, there is actually a simpler method, which is to directly falsify it.

The indicator is still the same indicator, the caliber is still the same caliber, and the adjective is still the same adjective, but the number reported to you is a false number .

This method has been demonstrated to us many times by Japanese companies in the past two years.

A quick search reveals many such cases:

1. In August 2017, Kobe Steel, a Japanese steel company, was exposed for tampering with product data and passing off inferior products as good ones. Potential victims include about 500 companies in Japan and abroad, covering many industries such as automobiles, railways, aviation, and aerospace.
2. In November 2017, Japanese carbon fiber material giant Toray Group admitted that its subsidiary Toray HC had tampered with the strength data of its automotive material products. The company's top management had long been aware of the matter, but argued that tampering with the data was not illegal and that "no safety issues were found."
3. In February 2018, TOKAN, a wholly-owned subsidiary of Mitsubishi Electric, admitted that 253 types of rubber products had not been quality inspected. Its products are widely used in equipment such as Shinkansen railway vehicles and elevators.
4. In February 2021, pharmaceutical company Kobayashi Chemical was exposed for fraud, with more than 200 patients suffering varying degrees of health damage after taking the company's drugs. Some patients lost consciousness after taking the drugs, causing 22 traffic accidents, and 2 patients died after taking the drugs. Even more outrageous is that the company's management admitted that they had known about this situation 16 years ago, but had left it alone.
5. In July 2021, Mitsubishi Electric was exposed for falsifying product inspection data, which may have lasted for more than 30 years. Mitsubishi Electric President Takeshi Sugiyama resigned.

This practice is very convenient and there is almost no loss, all you need to do is bow and apologize.

In this regard, Xiaorizi's data-driven technology is far ahead.

V. Summary

Today’s article is purely for fun , and the content is not truly “data-driven”, so don’t imitate it.

If you really have to do it, you can still add attributes, change logic, and change indicators. After all, you probably do it often. Sometimes the business does need some flexibility.

But you must never do things like falsifying data. Any slightly formal company will dismiss employees if they find out about subjective falsification , and there may even be legal risks.

Finally, 2024 is here. I wish all data analysts will not encounter such strange demands as the above in the new year.

Author: Fang Chaijun

WeChat public account: Sanyuan Variance (ID: sanyuanfangcha)