Data-driven business: Are you the driver or the donkey pulling the cart?

Driving force is a business concept that our market is discussing. When data becomes a factor of production, data-driven business also comes into being. What exactly is data-driven business and what is its role?

"Data-driven business" is a very fashionable phrase, but it is also a phrase that gives many students a headache. Because this thing can be seen but not touched. Every day, we just hear "drive" and "drive", but we don't know how to drive it.

Moreover, when I ask a question, the salesperson scolds me. When I give an opinion, the salesperson doesn't listen. They always complain that the data analysis is not in-depth enough or the suggestions are not specific enough. What should I do?

1. What is a driver?

When it comes to driving, the most typical scene is the driver driving a horse-drawn carriage. The horse pulls the carriage, and the driver holds a whip. When he shouts "Gee!", the horse pulls the carriage and runs fast. So the question is: how does the driver drive the horse to pull the carriage?
Think for a second -
The most intuitive thing that comes to mind is:

• Point out the direction: pull the horse to the right path.
• Increase speed: whip the horse on its buttocks.
• Increase supplies: Give the horse more food.
• Warning of danger: Pull the horse if you see a pit.

This is what drives intuitiveness.
These four points actually correspond to four major issues:

• Target question: starting point and end point.
• Method question: How to do better.
• Investment issue: how much resources to invest.
• Restriction issues: Will something go wrong under the influence of restriction conditions?

The same is true in enterprises. If you want to drive business, you must first solve four major problems (as shown below):

2. The role of data in driving

However, please note that these seemingly simple actions are not easy to do:

• Point out the direction: means to know the starting point and the end point.
• Increase speed: You need to know how fast the horse can run. Otherwise, if you just whip the horse, it will be in so much pain that it will not be able to run.
• Increase the supply: You need to know how much grass the horse needs to run for a day and how much it should eat at one meal. If the horse is too full, it will feel bloated; if the horse is too hungry, it will have no strength.
• Anticipate danger: You need to know where the potholes are and what kind of bumps the carriage can withstand.

As you can see, all four questions are related to data! The most important one is the quantitative goal. The starting point and the end point directly determine which road to take, how many roads there are, how much fodder to prepare on the road, and what pitfalls there will be.

So if the goal is not clear, nothing can be done. The goal must not only be determined, but also clear. Otherwise, just saying: "Take me around Ruoqiang County" sounds simple. But when you get there, you will find that this county is twice as big as Zhejiang Province! You can never run all the way to the end.

Here comes a question: What if we have never driven a horse-drawn carriage? What if we have never walked this road? What if we change the horses we use from Mongolian horses to Arabian horses? These are all new problems with no data. In the absence of data, we can do a test first. Run a small area to see the road conditions, see the horse's physical strength, and learn how to pull a horse-drawn carriage.

These are the two basic data-driven models:

1. For businesses that you are already familiar with, summarize experience, find problems, and eliminate bugs
2. For the newly launched business, test the effect, compare the groups, and find the rules

3. Data-driven advanced gameplay

Note: The above discussion is based on the premise that "the horse cannot be changed, and the cart cannot be changed." If the driver's driving goal is not "driving the horse" but "driving the cart to deliver the goods to the destination faster and cheaper," there are more things that can be done:

• Choose the right horse: which horse is faster and has better endurance?
• Change the car: oil the axle and strengthen the body
• Allocate cargo: Make several trips at a time, don’t tire yourself out in one trip.
• Use cheap ones: If the cargo owner can’t afford the money, then replace it with a donkey, which is cheaper.

This is a more advanced way of driving. Starting from the goal, not limited to the means, choose a more appropriate method to achieve the goal.

4. A stupid way to drive data

Question: If one day you see a rickshaw driver:

• Roll up your sleeves and pull the cart yourself, who needs a horse!
• Put a pig on the cart and whip it to pull the cart.
• Put a dead pig on the car and then study how to bring the dead pig back to life

Please tell me: Do you think this driver has a powerful cosmic force, or do you think this driver is a lunatic? Of course he is a lunatic! A carriage, a carriage, a carriage with horses is called a carriage.

Isn't it funny that the driver doesn't study how to control the horse or how to strengthen the body of the carriage, but tries to run instead of the horse! It's common sense that pigs can't pull a carriage, isn't it funny to use pigs to pull a carriage! Isn't it even funnier to try to bring a dead pig back to life! He has the ability to bring the dead back to life, so why is he still a coachman? ? ! !

However: When the driver becomes a data analyst, the carriage becomes operations, marketing, planning... and the whip becomes "big data", "artificial intelligence", "algorithm model"...

You will find that trying to whip an old sow to pull a cart is happening in various companies, and it is happening one after another and very exciting.

When many businesses hear the words "data-driven", they immediately give up and rely on data for everything.

Many businesses place all their hopes on "100% accurate predictions" without considering other possibilities or making contingency plans.

Many business people talk a lot about "data analysis must be specific", and then even leave it to the data to make a few pages and draw a few buttons.

Many businesses have problems with their own processes and low product quality, but are too lazy to make improvements themselves and so rely entirely on "accurate data push."

This is not asking a driver to pull a cart. This is not using a whip to beat a dead pig. Not to mention those who don’t even collect basic data, are too lazy to do data analysis, and don’t even know the business goals.

The kind who comes up and shouts: "We want to improve our performance, please analyze the specific ways to improve performance" - let's put it this way, when you go to the temple to worship the Buddha, you have to tell the Buddha a specific thing: "Please bless me to make 1 million" or "Please bless me to have a son" and so on. Besides, your data analyst is not a Buddha.

5. Another stupid way to use data-driven

Question: If one day you see a rickshaw driver:

• He said to the horse's long ears: "How do you think we should run better?"
• Driving a carriage? Driving a carriage is nothing more than shouting: "Gee!" "Gee!", and the horse will run away as soon as you shout.
• Ask in the WeChat group: Are there any senior drivers of Tou, Teng, or A? If you have any questions, you can pay!
• Search online for the PDF version of "21 Days to Master Driving a Bus from 0 Basics"

Do you think this driver is a scientist or a nerd? Of course he is a nerd! Only basic science has systematic books for people to learn. For something as practical as pulling a cart, if you don’t go to the scene to observe, what’s the point of finding Tou Teng’a? Because Xiao Ma Ge’s name has the character “Ma” in it, so all his employees can drive a horse-drawn carriage?

Many new data professionals often make the mistake of being limited to books, attempting to mechanically apply book knowledge to solve practical problems, and blindly believing in so-called "big companies" and "senior levels".

However, when you mention understanding the business, many newcomers go to the other extreme: they ask questions directly. They believe whatever others say. Where is the data? Where is the test? Where is the summary? They forget everything. Afterwards, they are still complacent: I have communicated with the business! Maybe they have been sold out and are still helping others count money.

Therefore, data-driven business requires the joint efforts of data and business. The business itself has a clear direction and strives to improve business capabilities such as design, operation, art, and product development. Only by carefully collecting data, monitoring the process, reviewing the results, summarizing experience, and testing innovations can the data be maximized.

However, the popularity of artificial intelligence in the past two years has made data-driven business even more chaotic. What is popular now is that businesses are too lazy to think and expect the magic of the "100% accurate prediction" model to help them see the sky.

Then the data thought that he could really make a "100% accurate prediction", as long as he could find a great god to give him a pdf version of "21 Days 0 Basics 100% Accurate Prediction" and save it to the computer D drive - dry goods - data mining - algorithm model folder. The result, of course, is that he is riding a blind horse and facing a deep pool in the middle of the night.

Author: Down-to-earth Teacher Chen

Source: WeChat public account "Down-to-earth Teacher Chen"