I’ve recently received a lot of questions from people who want to change careers in data analytics: Are data analytics jobs really scarce?

Actually formally set up in 2003 “data analysts’ professional recognition, data analysis jobs gradually hot up, made each big enterprise recruitment needs, many people feel data analysts so tight, scarce, head to also want to squeeze into the industry, even at the zero base turned, but the fact is not.

Take myself as an example. Ten years ago, I started my career in data analysis from e-commerce and basically witnessed the development of the whole data analysis industry. My conclusion is that data analysis talents are really scarce.

Some time ago, a friend of mine had a statement that I agree with. Don’t always mention how short the position of data analysis is, in fact, enterprises do not need so many data analysts. As you can see from the job postings below, data analytics is becoming a universal skill, needed in both product operations and project management, and real data analyst positions are rare.

Why is data analysis talent scarce?

For large e-commerce, retail and service service companies, there will be a demand for data analysts, while for small and medium-sized e-commerce enterprises, most of the basic data statistics and analysis studios are both operational.

As far as I know, every enterprise is doing data analysis, but it does not necessarily set up the position of data analyst. The data analysis work is carried on by the traditional position of sales, finance, secretary and so on. Therefore, there is a shortage of data analyst talents, and the position of data analyst is not in a shortage state accurately, but the poor supply and demand information in the market makes it difficult for enterprises to recruit suitable talents.

Data analysts generally need two to three years of work experience to be competent, and data analysts with more than two years of higher starting salary, small and medium-sized enterprises can not afford to hire, not to mention data analysts with five or eight years of work experience.

Many small companies choose fresh graduates to cultivate, but also just do data statistics work, no one can lead the way in the enterprise, so the data analysis post did not give the enterprise to bring due expectations.

In addition, data analysis talents are mainly concentrated in banking, finance, telecommunications, medical and other industries, which have higher requirements for the years of data analysts. Data analysts in the market are roughly divided into three types:

• Handyman analyst: Supply exceeds demand. The job content of the handyman analyst is to write SQL and take numbers, and make data statistics and simple reports according to the demand. Low barriers to entry and relatively easy work have led to an influx of people, but there aren’t that many jobs.
• Business Analyst: Supply and demand are basically in balance. The business analyst’s job is to evaluate the performance of operations, product modules and make simple recommendations. In addition to data analysis skills, the entry threshold also requires business, operation and product knowledge.
• Business Analyst: A lot in short supply. The job of a business analyst is to make suggestions on future operation and product direction by analyzing and judging the current situation of the market, competition and enterprise, so as to increase the share and increase the revenue. Entry threshold in addition to these above, but also to understand business, understand the market, understand management.

What do you need to know about career change data analysis?

First of all, we should understand that data analysis skills have become a universal skill. Whether you want to be a data analyst or not, you should have them, which is very helpful in the product, operation and other positions of enterprises.

Secondly, data analysis focuses on analysis. Data collection and data processing in the early stage can be learned by hard learning, but the ability to analyze data can only be supported by methodology and experience. The test is your data thinking and insight. No matter how good your data chart is and how sophisticated your data processing skills are, if you can’t analyze and interpret the data, it’s completely worthless.

Finally, do your own career planning, how to do? The best way to do this is to understand the business model of your industry, understand the real relationship between traffic data, and understand how business models are represented through data disassembly.

Therefore, it is more important to emphasize that data analysis is not just to display beautiful things or prove oneself. It is meaningful to use data analysis to influence decisions, whether it is product operation decisions, enterprise development decisions or personal career decisions.

For example, your boss suddenly asks you how the company’s business figures are in the last three months. You say how much revenue and traffic have increased in the last three months. You think your boss really doesn’t know? He’s looking at the core data every day. What are the answers he’s looking for, where are the optimizations in the way traffic is captured, where are the areas of revenue growth that he can continue to exploit. Not only do you need to be able to process data and analyze data, but you also need to be able to understand data in multiple dimensions.

What do you need to prepare for career change data analysis?

To put it simply, there are several hard requirements for data analysis at present: education, profession, experience and skills

Data analysis is not tall to formal schooling major requirement really, junior college can, major is more at present unlimited, you learn veterinarian but self-study very good I feel OK also. Experience is critical, but some entry-level positions are open to career changers. The skills that should set the bar are actually very low.

How do you prepare for a career change?

1. Knowledge of statistics: percentile, boxplot, standard deviation, Pearson correlation coefficient, Bayes’ theorem, normal distribution, Chi-square distribution, hypothesis testing, etc.

2. Preliminary understanding of operational data indicators channel conversion rate: PV, UV, retention rate, turnover rate, repurchase rate, GMV, etc.; understanding of Website analysis tools such as Google Analytics and Baidu Index

3, familiar with SQL syntax: such as Mysql database

4. Familiar with common python data analysis libraries: Pandas, Numpy, Matplotlib and Seabron, able to use Python for data visualization, data exploration and data preprocessing

5. Familiar with Office software: proficient in Excel, common functions and PivotTable, etc

6. Have a preliminary understanding of common machine learning models: decision tree, RF, clustering, etc., and be able to deduce the least square method by hand

7. Understand common Linux commands

8. Know text mining: regular expressions and word clouds in Python

conclusion

In fact, skills are only one hand, learn SQL, do a report can also find a number of points of work. SQL, Python, machine learning have learned, can also find a number of points of work, perhaps the treatment is similar. However, I think in the current era of fierce competition, a sufficient preparation for career change can increase your chances of getting a high-quality factory, and the improvement of tools may also give you a broader vision.