Recently, together with Yuri Trostin, Head of Analytics and Data Science at Worki, we held a webinar "How to change your profession and become a cool analyst?" Watch the recording of this online intensive
Why do you need analytics?
The goal of any business is to make a profit. Profit is revenue minus costs. Profits must be maximized over the long term. How does analytics help here?
There are three major areas in which analytics adds value:
1. Formation of business processes.
Where is the business heading, in the right direction, and how are key customer and business segments feeling? With this data, the analyst can suggest solutions to certain problems. Let's say something goes wrong in business, and in a rapidly accelerating world, time is really the most valuable resource. You need to solve problems and bring new hypotheses to market faster than competitors, and this is where your analyst can help you.
2. Make informed decisions based on data.
Here is not only about alerting, but also about the formation of key decisions. Doing any business is accompanied by uncertainty, and you cannot get the full information. If the information was always complete, accessible and evenly distributed among market participants, then analytics as such would not be needed. But business also would not have super-profits. If you have the relevant data for making a decision and an understanding of how to extract information from it, then automatically you win over those who do not have the data and this understanding. Of course, in business you cannot always make the right decisions, but, for example, using a data-driven approach, you can minimize the share of bad decisions. Of course, for this you need specialists who can analyze this data for you so that it is complete and consistent, that is, consistent.Then, in the long term, you can benefit from it.
3. New ideas for business, development and experimentation.
At its core, this is primarily some kind of idea, and it can either enter the market or not. The more relevant ideas and the more experiments are generated, the more profit you can get in the future. The beauty of IT products is that by examining the patterns of behavior within the service, you can understand insights that would seem initially not obvious, but which speak about how the user will be better off through analytics and data. In addition, the accumulation of data about what you viewed in order to generate more relevant content for you lies in the same direction. For example, this is done by YouTube, Netflix, VKontakte and other companies where the recommendation system is very widely developed.
The key question for any analysis, during and after it:so what?
What does all this mean to the business? Are you somehow improving your understanding of what is happening in the business at the moment? Are you generating more ideas and experiments through analysis? Are you making better and more recent decisions?
If at least one answer to these three questions is "yes", then the analyst is not doing his job in vain. Analytics is not just numbers and numbers, it is a powerful tool that allows you to run a quality business. Companies that understand this are ready to seriously invest in analytics, because they know that despite the cost of the process, much more can be gained from it.
What is important to look for when looking for your first job?
If you have decided for yourself that analytics is what you need, then when looking for your first job, you need to find a place where you can best pump. By pumping, we mean not only hard skills and the use of tools, but also how to approach problems, how and what data to use.
What should you look for when looking for your first job?
The points will go in descending order of importance, from most important to least:
1. The most important thing is the team and the leader.
These are the people with whom you will be studying for the next six months, a year or two. Before boarding, ask yourself: do you want to learn from them, do they inspire you, are they cool at what they do?
Try to find out as much as possible about them: look at their speeches on the Internet, if they have any blogs or channels, maybe they write texts in specialized communities. If at least one person in the team does any of this, then this is a good sign.
In the interview, ask in detail what is expected from June, what is the format of interaction in the team in the company. Remember that your main goal is to level up and get out of there with a baggage full of knowledge and experience.
2. The company itself.
This is not about the office and working conditions - this, of course, is not bad, but this is not a long-term motivation. This is about the message that the company itself broadcasts.
High-quality personal growth can only be combined with motivation. If there is no motivation, then you cannot pump cool. If your vibe resonates with the company vibe, only then can you effectively improve your skills. It is better to immediately find yourself some place to your liking. Analytics is now needed everywhere: in e-commerce, classified, foodtech, gametech, HR, media, logistics, etc. In reality, there is data everywhere and you need to work with it in order to effectively manage your business.
3. A stack of technologies that the company uses.
As June, you may not understand them, by and large no one expects this from you, but if the guys use Excel and they have the same MySQL database, then you should be wary. Yes, Excel is a super powerful tool, but if a company has a great data stack, it means that it understands its importance and is ready to invest in it. And, most likely, the business has a great team, which means that you can pump better.
What can you suggest when looking for your first job as an analyst?
1. Knowledge of SQL.
If you can't get the data you want, then you can't do the analysis. You can get the data using SQL. Yuri Trostin had a lot of rejects due to the fact that he did not know SQL. Then, of course, he had to learn it.
SQL is different:
- SQL, 80- . . SQL sql-ex.ru. SQL, .
- IT- SQL, ClickHouse. ClickHouse — , , .
ClickHouse is now used everywhere by everyone, for example, Mail.ru Group, Avito, Yandex. Its syntax does not differ much from the main one, although, of course, there are differences that make it more functional when working with ClickHouse. Its tasks are focused specifically on analysis, on leads, and not just on data extraction.
2. Python.
This is an industry standard for data analysis, data science, and you can also create visualizations right away. Knowing Python allows you to perform certain operations much faster if you use it in conjunction with SQL, as opposed to when you just have SQL. Knowledge of Python will be a super plus for a potential junior.
Yuri Trostin notes that Python was much easier for him than SQL. He hung out a lot on kaggle.com, did competitions there. There are also many different scripts for analysis, cleaning, data visualization in Python. The second point is the courses. For example, the same course from ProductStar.
3. Data visualization systems / BI-systems.
Without data visualization, your analysis makes no sense. With the help of a BI system, you can analyze data, visualize it, collect charts into a single dashboard, which will give a better idea of what is happening in business, and you can also generate insights from this data on the fly. Products like Tableau, Power BI, QlikView, these are all related to BI functionality. They are similar to each other, so if you get acquainted with any one of these products, then it will not be difficult for you to switch to something else later.
4. Specific products used in analytics.
It's no secret that there are marketing analytics, business analytics, product analytics. Some places have strategic analytics, call center analytics, support lines, etc.
Narrow tools for marketing analytics are Google Analytics and Yandex.Metrica. For product analytics - Amplitude, which is needed to analyze user behavior in applications.
5. Econometrics, A / B testing, Data science.
At the junior level, this is not so important, but in the future you will definitely need knowledge of such tools if you want to excel in analytics.
You don't need to jump to the tools below if you haven't learned the tools above first. If you don't know SQL yet, then you shouldn't start learning Python, etc.
What else can help you when looking for your first job?
1. Solution of business cases.
When solving business cases, you will learn to think and speak in a structured, fast and clear manner. You will learn to formulate hypotheses, query relevant data, conduct qualitative analysis and draw correct conclusions. It will also teach you to clearly convey information about the work done to different people.
2. An understanding of how IT works.
This is because the analyst is usually between the business and the technical team. An analyst needs to be able to find a connection with both the business and the technical team.
3. Dating in the industry.
This is especially valuable for those without a technical background. Through dating, you can get recommendations, this, of course, is not determinative, but at some point it can help.
There are various offline and online meetups where you can meet people. There is also a large Slack community, Open Data Science (ODS), with over 30 thousand people, among whom you can also find people of your interest.
What to expect at the interview?
Typical interview scheme:
1. Acquaintance;
2. The technical part.
Checking the skills you listed on your resume;
3. Homework.
You are either provided with a data scheme, according to which you should eventually send not a specific answer in the form of numbers, but SQL. Alternatively, you will be given a data set that will need to be analyzed using Python or another programming language and then send recommendations for this analysis;
4. Motivational interview.
Find out why you need this particular job.
What should Jun do at his first job?
The main thing in the first job is maximum pumping.
1. Communicate with the team as much as possible.
Always consult with your colleagues when solving any problems, so you will absorb their experience and do better work.
2. Try to understand exactly how the business works.
Ask questions:
- What is your company selling?
- What's the economics of one sale?
- What is the monetization model?
- What does a user get when they use your company's product?
This will help you form a big picture of your business, while also helping you analyze the data and formulate hypotheses.
3. Communicate not only with your team.
Chat with all kinds of people inside: development, product, marketing, sales. They can share cool business and market insights with you.
4. Expand your areas of expertise, not just analytics.
5. Do not sit in one place.
When you realize that your responsibilities are beginning to repeat, you cannot take from this place as much as you took before, then think about it, perhaps you need to open up to the offers that come to you so that you can continue to grow as an analyst.
What to look for when forming a team?
1. Desire and passion to work with data.
If you like to look for patterns in data, if you understand that there is a physical meaning behind the data, if you can influence something with the help of data, then this work is definitely for you.
2. Drive.
It's about the desire to change things. A great analyst should be proactive.
3. Diverse experiences.
When a person has a wide range of experiences, they can add their own non-standard point of view to the problem. It is more interesting to work with such people.
4. Motivation.
It is important that the person is clearly aware of:
- What does he want to get as a result of this work?
- Why did the person decide to work with data?
5. Technical skills.
Nowhere without them.
Useful links from Yuri Trostin:
Victor Cheng is a consulting icon ... He has cool books and lectures on YouTube, as well as audio recordings of case interviews. You can find and listen to them.
A very common book in the field of consulting, which is a collection of business cases. Try to read and solve the cases from this book yourself.
The book is intended for a large layer of specialists, from Juns to pros. The book develops well the idea of why analytics is needed at all. She even creates a framework for how to think about analytics in a company.
Yuri's favorite YouTube channel. There are many quick courses, including Computer Science courses. With this course, you will be able to learn the basics of Computer Science and understand where it originated.
A good book on how the Internet works. Simple enough, I recommend reading it.
This source allows you to take a fresh look at analytics, the importance of individual metrics. This guide contains tutorials on how you can use all this knowledge in Amplitude.
Yuri Trostin's speech in Minsk, where he talks about how they make data-driven startup Worki.