Ultimate guide to a data analyst role: skills and requirements

In eight years of work in analytics, I have interviewed and hired hundreds of people and have a good idea of the ins and outs of the analyst market.

The key knowledge here is that this market practically doesn’t exist. In 2019, I hired 34 analysts for my team, 23 of whom (68%) were interns or juniors. I would have been happy to hire someone more experienced, but people of such level didn’t exist, so I had to hire people with potential and help them grow.

In comparison, we hired 23% junior team members (five people out of 22) for data engineering teams, so the market is there. Data engineering is common and well developed in banks, telecom, and retail, which means that there are more ready-made specialists in the market.

This essay serves two purposes.

First is to clarify the terms in which we think about the levels of analysts. This will reduce the existing entropy in the market, where an arbitrary set of expectations and skills can be hidden behind a job opening or an analyst’s CV, ranging from project management and systems analysis to automation of routine business operations. In this market environment such prefixes as junior/senior/leading carry no information at all.

The second purpose is to provide a clear roadmap for growth and development as a data analyst or a person who has to do the work of a data analyst but has a different title to make it more applicable to everyone. At Yandex.Taxi, we are forced to build a growth career ladder for our employees, because otherwise we won’t be able to cope with the demand. The very formalization of analysts’ levels described in this essay is a consequence of this approach. However, not everyone works in large companies, and not everyone has access to a strong mentor. So this essay aims to help such people take a look at their growth points and work on them.

→ Test your product management and data skills with this free Growth Skills Assessment Test.

→ Learn data-driven product management in Simulator by GoPractice.

→ Learn growth and realize the maximum potential of your product in Product Growth Simulator.

→ Learn to apply generative AI to create products and automate processes in Generative AI for Product Managers – Mini Simulator.

→ Learn AI/ML through practice by completing four projects around the most common AI problems in AI/ML Simulator for Product Managers.

What does a data analyst position imply

A data analyst is a person who helps the team with the following:

• Make decisions more objectively based on facts and data (as opposed to opinion, intuition and experience).
• Look for growth opportunities for the product and business.

An analyst’s work involves working with data with the help of programming languages (SQL, Python, etc.), creating dashboards, and automating processes. But these are just tools to achieve the two goals described above. If a person simply does these tasks, then she should not be considered an analyst. Perhaps she is a data engineer or an RPA developer, but these are completely different roles with a different set of requirements.

The analyst concentrates on exploring data, structuring complex systems, and understanding processes to benefit the business. The analyst answers questions , creates thinking models and frameworks, and develops recommendations derived from them, which lead to an increase in business performance.

In most cases, the way to achieve this is by working with data. But working with data is not an absolute necessity. Analytics helps you make decisions and plan actions in the product and business. In some cases, you can make a decision without analyzing the data, by simply formalizing all possible situations and decision paths, and then discarding most of the options and relying on what the team already knows.

Basic data analyst skills

In order to effectively cope with the described tasks, the analyst needs to:

• have excellent mathematical knowledge: no one wants to check formulas in the analyst’s work
• understand the basics of probability and statistics: the analyst must be able to test hypotheses, understand different types of errors, dependence/independence of random variables, etc.
• be able to think abstractly and mathematically: if an analyst uses a method or an algorithm, she must know the area of ​​its applicability
• possess critical thinking skills: the analyst must be able to avoid the traps of cognitive biases
• have a product mindset: the analyst must be able to express user experience in metrics, but must also be able to go beyond the metrics and examine the users who are trying to solve a specific problem
• have a business mindset: the analyst must be able to digitize the company’s business processes and market changes, and link them together with the product and users
• be a techie: an analyst doesn’t need to be a highly skilled developer (i.e., write efficient, fault-tolerant, and scalable code), but should have no technical blockers to solve business problems. The analyst should have no problem reading technical documentation, pulling data from a new database, writing a parser, or using some APIs for automation, etc.

An important property of a good analyst is self objectivity. The analyst must control her desire to appear better than she really is. Even the most powerful analysts make mistakes—this is normal and there is no development without mistakes. So, it is very important to be able to track one’s mistakes admin them and quickly communicate them to the team, especially if this can change a previous decision. Concealing errors is a red flag of unprofessional behavior by an analyst.

In Yandex.Taxi company, analysts are assigned to teams, but there is no strict attachment to the subject area, be it marketing, product, or operating system. The requirements for the level of thinking described above allow analysts to switch (and of course take time to adapt with new tasks).

Summing up analyst levels: from intern to super stars and super bosses

First, you should be careful when using this table. In Yandex, the understanding of grades comes as a result of calibrations in the performance review process. In this process, analysts’ team leaders explain the assessments of their team members based on their performance in the past six months. Usually, you need to go through at least a couple of reviews to understand the requirements well enough to broadcast them—this is part of the culture, and it is not absorbed instantly.

The spreadsheet is a cheat sheet: it helps a lot if you know what it is about, but it may be misleading if it is your only source of information. Therefore, the table should be used only along with the text below.

Further, we will consider two branches of analyst development: as a specialist/individual contributor and as an analysts’ manager.

You can see the full version of the table with the data analysts positions here.

Data Analyst Levels (Specialists/Individual Contributors)

1. Intern data analyst

Above, I described the basic skills for an analyst role. Consider them as screening requirements, which means even intern-level analysts must have the foundations of those skills. Most often, the intern analyst doesn’t have any previous special experience besides university education. But she has the necessary qualities to grow into a strong analyst in the future.

And then you may wonder: aren’t these requirements a little too much? The answer is no, they’re not. And here’s why.

An intern will take up the energy and time of the leader and colleagues (instead of becoming an added resource, as it might seem) for the entire three months of her internship. Therefore, it makes sense to hire only people with noticeable potential as interns to maximize the return on the time and energy invested into their training.

The most important criteria for hiring a intern analyst are academic background, mindset, and university or pet-projects programming experience.

Academic base

Analysts work with statistical data. The intern may not have practical experience, but during the internship, she will figure out why she took the Statistics course at the university. If there was no course, the volume of training for the intern becomes too much for a company to handle. At my company, we probably wouldn’t take this person on.

Mindset

Formalizing mindsets is very difficult. But we are trying to define some tests and questions that will help us see the way the person thinks about the world through data that doesn’t depend on the candidate’s working experience. For example, we might ask a candidate to take the process that people living in cities face every day and describe it with the help of product metrics.

Thus we test a person’s ability to put herself in the shoes of a consumer and highlight the most important thing in the consumer experience. This is not something you can explicitly teach. It is more of a way of looking at the world, a level of empathy, and the ability to develop a critical perspective.

A few examples of tasks we give: come up with custom metrics for a traffic light, digitize the process of heating water in an electric kettle, etc.

Programming experience

It’s not about professional-level programming. As a data analyst, you need to be able to read documentation, quickly understand and use data-manipulation tools, and automate your routines. For instance you should be able to find your way around implementations of SQL, work with different

Python libraries (data processing libraries such as pandas, visualization libraries such as matplotlib, etc.), and have the ability to explore and use APIs.
In general, techies learn programming much earlier than they start real jobs, and they grow technical skills easier and faster. This is what we want from our job-seekers.

Soft skills

Without self-discipline, responsibility, and the ability to communicate constructively with colleagues, it will be difficult to achieve any result in a team. There are exceptions to this rule, but they are very rare.

Setting goals

The intern analyst is given tasks that are clearly formulated and well-formalized. The tasks for the interns are set by the senior fellow analyst (mentor or team leader). The senior also checks the results of these tasks before passing on the data. At this point, it is still the manager’s responsibility to check and comprehend the data, visualize it, and communicate the conclusions. But the intern can make her first approach to these parts of the work.

2. Junior data analyst

Data processing

A junior analyst is an intern who has mastered data processing tools. The junior analyst has no restrictions in the tasks of transforming the available data to the required form. Her work might have some flaws—maybe unnecessarily heating the computing cluster with inefficient calculations or spending a lot of time on simple tasks—but she will solve the problem in the end.

Data distrust / data validation

The junior analyst has enough experience to know how far to trust the data. When a junior analyst starts working with data, she learns the nature of the data and checks it, making sure that the data is exactly what is expected to be. For example, she might check to make sure the indicators are in the required dimension, the distribution of the values ​​looks reasonable, there are no strange outliers, the data reflect the real picture of the phenomenon under study, etc.

Setting goals

The junior analyst lacks experience in a real product and business, so she tends to solve problems in the form as they come to her from the colleagues:
-Do you need to insert X into this Excel spreadsheet? – Here it is,
– Do you need to make a dashboard? – Draw on a piece of paper what it should look like and I’ll get it ready.
etc.

As a rule, when setting a task for a junior analyst, a detailed algorithm is discussed with a description of the data that needs to be used, the method of transforming this data (filtering, grouping, joins) and a literal description of what the result should look like (should it be a graph, what should the axes be, normalization, signatures, visualization method).

A business customer can directly give tasks to a junior analyst, but I don’t recommend it. Ideally, all tasks should still be supervised by a mentor.

As her experience grows, the junior analyst can work more confidently with familiar data and with less-detailed task descriptions from the mentor. In this case, the key factor that determines the analyst’s level is her proficiency to work with unfamiliar data.

Implementation and application of work results

In general, junior analysts can’t prepare analytical reports or research in a way that provides clearly justified recommendations. Moreover, she can’t mature her recommendations and conclusions to real changes at product, process, or business levels.

Junior analysts are still of little concrete value to the business, so it is critical that they grow as quickly as possible. The principle that boosts the development of a junior analyst is “do as I do.” A more experienced analyst (in the presence of his mentee) discusses the new problem with the project team, asks clarifying questions, and plunges into the context of the problem. This dialog results in a specific approach to solve the problem. Later on, a senior team member breaks down this approach into data-processing tasks that the junior analyst can handle.
This allows the junior analyst to observe how business problems turn into coding and graphing tasks. Over time, the junior analyst should learn to do this on her own.

Why it is important to have an in-depth discussion of the task with junior analysts

When leading a project, there are questions and problems that analysts can help with. Usually, when business customers go to an analyst with a problem or idea, they also have an approximate solution in mind (“can you build this graph for me?”). But a project manager, designer, or product manager may simply not have enough information about what other data and tools are available and needed to get an answer. Perhaps there is a more accurate or simpler way to solve the problem. Or maybe, within the context of the hypothesis that we have, it is not necessary to process the data at all. Maybe it is enough to look at the dashboard, where there will be a graph that allows, with some assumptions, to answer the question asked (without being too concrete or ideal).

3. Analyst Level 1 (Middle data analyst 1 step)

Analyst 1 has all the skills of a junior data analyst in terms of technical work with data and critical thinking, but at the same time moves further in the following aspects of her work and business impact.

Setting goals

This is the first level when the analyst becomes independent. Almost all the tasks she performs are in direct interaction with the team or business customer.

The junior analyst moves to this level because she wishes to have more influence on the business or product. It is clear that wishing alone is not enough. Before being qualified for this opportunity, she needs to stack up enough experience.

Analyst 1 is better positioned to get to the bottom of the ideas and questions brought to her. She starts thinking about tasks in terms of decisions to be made, not in terms of working with data.

The complexity of the tasks and the depth of the solution

The analyst at this level handles most of the simple tasks on her own. However, given the little experience she has in solving problems in a business setting (data processing should be a fine starting point for a junior level, as we remember), an analyst of this grade may lack depth of thinking or breadth of context to better understand the nature of complex problems and select the best solution.

In general the level 1 analyst will struggle with the following:

• situations with a high degree of uncertainty
• tasks around complex and multi-layered business processes
• tasks with a demanding customer, when you need to carefully work with objections

In these situations, analyst 1 needs the help of a senior colleague to understand the problem, decompose it, or present the results (in a way that can be applied).

Analysts 1 are more likely to suffer Dunning-Kruger syndrome, a state where a low-skilled person overestimates her abilities. She has already reached a stage where she has become independent and has learned how to help businesses. It seems to her that she can handle any task (and in general she has grown to a senior analyst by this point).

The problem is that Analyst 1 doesn’t understand which of her tasks she could dig deeper into. Therefore, her mentor tries to stay on top of almost all of her tasks and the solutions she chooses. The format of the control can be different, for example: stand-ups of the whole team or regular 1-on-1 meetings.

Analyst 1 copes well (and independently) with tasks that have a clear solution, such as preparing and analyzing A/B tests. If an analyst demonstrates a proactive position and not only answers the direct question of the A/B test, which is “can we release it or not?”, but also looks for (and finds!) issues that can improve the performance of the product, this initiative stands for a higher grade.

Analyst 1 can do complex research only under the supervision of a mentor who helps with the essence of the task and the decomposition, as well as the design and presentation of the result.

Soft Skills

A high degree of independence means that an employee has the skills to manage time, form expectations, and predict deadlines. Analyst 1 is already responsible for the final result.

Data

If analyst 1 doesn’t have enough data to solve the problem at hand, she negotiates with the team who can log or supply it. It may be that she doesn’t do it optimally, but she can already solve some local issues. This skill is polished in the next grade.

4. Analyst Level 2 (Middle data analyst – 2 step)

Analyst 2 is an autonomous analytical unit who creates real value for the business. This is a more experienced version of the previous grade, but there are also some significant differences here, such as a better understanding of the context and an increased level of reflection. Below I will talk about why this is important and how the nature of the interaction and the benefits of this analytical unit is changing.

Context and proactivity

Analyst 2 has a deep contextual grasp of what’s going on in her product or part of business. This allows her to be proactive, come up with ideas and suggestions, and help the team without being asked to. This proactive role dramatically increases her value to her team and organization.

Why it is important to be in context

The most vexing problems are the ones that are difficult to formulate. These are situations in which there is some kind of pain or need, but there is no clear solution. There are no tickets created for such tasks, which means that to help solve them, you need to be proactive. But if an analyst has no idea about the problem, she won’t be able to help.

You cannot be a high-level analyst and be out of context of the problems and tasks of your business and product.

Self-awareness

An important difference between Analyst 2 and Analyst 1 is the increased level of self-awareness. To become a good analyst, you need to question everything and critically rethink all the things coming your way. I can say that science is based on the same approach and successful analysts, like decent scientists, possess this quality, too.

Because of her high level of self-awareness, analyst 2 is not afraid to be left alone with the tasks or some branch of business. You can be sure that no nonsense will happen there if she is in charge. If necessary, Analyst 2 will be the one to ask for advice.

As a result, we get a special trait when managing Analyst 2. Since Analyst 2 will come to her manager for advice if necessary, the manager doesn’t have to closely track her daily routine and they can discuss higher-level things during their meetings.

The complexity of the tasks being solved and the depth of the solution

Another important trait of analyst 2 is that she thinks of the business at a higher level of abstraction. Analyst 2 should be able to see some systematic directions of development in the part of the business entrusted to her during routine operational tasks. She must understand how tasks relate to goals, and where and why the product/company is moving globally.

Analyst 2 is perfectly knowledgeable in the subject area, which allows her to dig deep. She knows how to start from business problems and reach the right metrics that need to be optimized. If asked about some issues, she knows how to go deep to the very core of the problem and knows how to transform its essence into data that can be presented and discussed.

Analyst 2 solves problems that are formulated in a business language/setting. She formulates the results of her work in the same business language. If this is an analytical report, it contains recommendations on how to proceed and what to pay attention to (i.e., points of growth and potential problems). If it’s a dashboard or a chart, it will help make decisions.

The team trusts the results that Analyst 2 brings to the table. This is a characteristic of both the accumulated experience and karma, and how the analyst at this level formalizes and communicates the results of her work. Usually, in her tasks, the methodology is clearly described, the scripts and code are there, the assumptions and the scope of the result are described well, and there is a unit with all the well-written conclusions in a clear and understandable wording. If the analyst’s results are not trusted, they won’t lead to any action, which means her work is useless.

Most of the results an analyst 2 brings should be actionable. Analyst 2 is an experienced professional that we have taught or hired to bring value to the company. The analyst’s main task, as we recall, lies in improving metrics, and not in writing SQL queries or generating graphical reports. Therefore, there must be a strong actionability bridge between data analysis and business outcome.

Data

Analyst 2 can be an independent customer of the Data Warehouse (DWH) team – describing exactly the datasets that she needs to work with.

Mentoring

Quite often Analysts 2 are assigned interns or junior analysts. Analyst 2 must be able to control and train junior colleagues. This means she must have the skill to delegate some simple tasks to them to help develop their expertise and the degree of independence.

5. Senior data analyst

This is a kind of superhero, and there are few of those. Even in everyday work, this person brings tremendous benefits to the business, expressed in the impact on the key product metrics or the business process entrusted to her. Compared to the previous level, she can solve problems that have a greater depth and complexity. She also has greater autonomy and proactivity. Unfortunately, not every analyst 2 will grow to this level.

The complexity of the tasks being solved and the depth of the solution

The senior analyst can solve problems that do not have a “head-on” solution—and the final result of her work will be credible. She knows how to communicate the result so that it is well understood and accepted. A typical example of the kind of tasks the senior analyst will tackle is evaluating a launch without using an A/B test, when nothing is clear to the naked eye (e.g., loyalty program, the influence of Drive (a service of a car leasing or a car sharing) on Taxi, cannibalization in performance marketing, etc.)

The senior analyst has the experience and context that allows her to anticipate whatever challenges might come up in the future.

A senior analyst is a valuable resource. From a company perspective, it is important to strive to ensure that every result of the senior analyst leads to positive change. If a senior analyst has found something valuable, she must make sure that everyone involved knows about it, pushing the necessary change in the product or business.

The senior analyst has great communication skills, she knows how to find a way toward different people in the company, she knows how to choose a communication method that suits the problem being solved, and she can explain complex concepts using simple language.

Data

The senior analyst’s experience qualifies her for being a customer for the DWH (Data Warehouse) team. She knows the requirements for the needed data and logging, allowing her to solve both current tasks and those that may arise in the future.

Mentoring

As a rule, the tasks and projects of a senior analyst are quite big. She can decompose them and distribute parts of them to less-experienced analysts in the team.

In a growing company, a senior analyst almost continuously has an intern or junior analyst to train by her side.

Management

As a rule, the senior analyst is driven by goals (“Now we are at point A, and we want to get to point B”), most of which she determines herself and coordinates with the management. Most often, the regular communication between the senior analyst and the manager takes the form of a strategy discussion, psychotherapy and solving HR-questions.

6. Lead data analyst

This is a true legend. These are crazy rare. We are now in a territory where there is no established market and requirements for such people.

The complexity of the tasks being solved and the depth of the solution

A lead analyst is an expert in her subject, which is a science-intensive field (e.g., dispatch algorithms or behavioral economics). This level requires a strong academic background, a broad outlook, and many years of experience. In general, this person is perfectly qualified to write essays in peer-reviewed science journals or speak at scientific conferences.

To implement projects in complex subject areas, you need to have strong communication and managerial skills, and take on educational and promotional functions.

Management

The lead analyst determines her priorities and ways to bring value, and only coordinates this vision with management. If there is a change of priorities, it usually happens by refocusing: “It seems to me that this issue has now acquired great importance for the company. Please think about what could be done here.”

Data Analyst Leadership Levels (Analyst manager)

The analyst manager’s job is to scale her value through the team. For example, if John is a senior analyst, then taking a manager’s position, John will do well in his team if the team of five analysts (including himself) will benefit at least as much as three Johns combined.

The manager’s grade is usually one level above the most senior team member. The trigger for the growth of a manager’s grade is the presence of traction in raising people to their current level and their effective management.

It’s great and logical when senior analysts become managers, but life is complicated and sometimes, when the company is growing rapidly, some mid-level specialists have to take on the management function.

Analytics Manager 0

In reality, there is no such grade. But I started with this category to emphasize that mentoring is not management, as some aspiring mentors might think.

Analytics leader 1 (Analyst team lead)

Team structure

This is a beginner leader. She has several analysts in direct horizontal reporting. The most appropriate number is up to five, but sometimes there are more.

Mentoring

Analytics Leader 1 has to raise the members of the team through mentoring. She helps them with things such as decomposing big tasks to the required level of complexity, participating in setting, solving, communicating results, teaching new approaches and technologies.

Motivation

Analytics Leader 1 motivates employees every day through regular meetings with the team and individuals. She formulates inspiring goals and helps the team achieve them. She makes the team feel as one, driving synergy from the interaction of different analysts, even if they are dealing with different tasks.

It is difficult for a beginner manager to deal with tasks such as bonuses, compensation, and counter-offers due to the lack of accumulated experience. Usually her superior helps her with this.

Expertise

Analyst Manager 1 has expertise in the part of the business for which she is responsible for. She is the key person shaping the data requirements for her part of the domain (e.g., performance marketing or product). This data will subsequently be reused by other teams. It is important that this data is correct, conveniently designed, and well documented. This allows you to lower the entry barrier for the rest of your colleagues.

Processes

Analytics leader 1 has to build interactions with business customers and suppliers so that her team and colleagues can work effectively. If she sorts these problems out correctly, her colleagues shouldn’t have questions such as who they should address their tasks to or how the task is set, how analysts can help them in general, how long will they have to wait for the result or how to influence the priority of their task. On the other hand, the team of analysts should be in the same context with the business, and they should understand each other perfectly.

This usually means that there are some interaction rules. They can be different in various teams and companies. To name a few, there can be analytics team stand-ups, regular meetings concerning priorities with the business, seminars where members exchange their experience, regular one-on-one meetings with the manager, analytical reports on the quarterly results, rules for working on tasks (e.g., tickets, version-controlled code, metrics’ titles in reports, etc.)

I would like to emphasize that while all of the above are the tools, the essence is this: analysts should bring maximum benefit and work effectively.

Team Leader Level 2 (Head of analytics)

A significant difference from the previous level is in the structure of the team and the increased ​​responsibility.

Team structure

Team Leader 2 has a significantly larger team, with intermediate leaders in it. Usually there are 3-4 of those. From the responsibility point of view, everything remains the same, but on a larger scale.

Processes

A leader at this level deals with systemic tasks that help her team and company work more efficiently. This may relate to the following:

• HR processes
• processes for evaluating results (performance review)
• analytical infrastructure issues
• reporting tools
• etc. (the current list is incomplete)

In general, a leader of this level is an experienced analyst. She knows very well what tasks analysts have and what difficulties they face in the process.

One of the areas of responsibility at this point is a convenient analytical infrastructure and environment, such as virtual machines, Jupyter hubs, libraries that facilitate routine, convenient data marts, etc. She is a qualified customer for the data warehouse (DWH) team.

An analyst leader may not be a machine learning professional, but is certainly familiar with the subject area and can speak the same language with a data scientist. She understands well when a problem should be tackled using machine learning methods. This allows her to be a qualified customer of the ML team.

If the company is small, then the DWH and Data Science teams can be subordinate to the head of analytics at this grade.

Leader 2 is where top management and analytics teams intersect. She maintains the context and a unified information scope in her team and across the entire company.

Mentoring

At this level she supervises level 1 leaders. This means that she knows how to influence teams and processes in them without any direct intervention, but through an intermediate leader instead. She serves as a mentor to aspiring leaders.

Analytics Leader Level 3 (Chief Data Officer / Chief Analytics Officer)

C-level *. This position belongs to the top management. She has excellent knowledge and understanding of the context (including non-public) in which the company and the market live, and is well aware of the goals and problems of the company. In a broad sense, this person is responsible for the performance of the company, helping to improve it through analytics and data manipulation. To be a successful director of analytics, you definitely need to be a pro in the analytics field. But describing the role of a director is more like describing the role of any other director in the company, aside from the subject area.

* Large corporations have rather complex and branched structures. This refers to the C-level at the level of an individual business unit or product, the head of which would be called CEO if it was a separate company.

Team structure

Dozens or even a more than a hundred people (depending on the size of the company itself) of different specialties:

• analysts
• machine learning specialists
• visualization and reporting specialists
• data engineers
• system analysts and project managers
• market developers and UX specialists
• (the list is incomplete)

Leadership

The director is the absolute authority and head of her direction. Analysts and engineers don’t want to just accomplish tasks. They want to influence the company, product, and customer experience.

The leader’s task is to inspire the team towards greater results, to show how the goals and objectives of concrete people are related to the company’s vision and goals.

Why is it so important at this grade? Of course, leadership is important at lower levels as well, but if it is lacking there, then more senior leaders can support you. But when you are the boss, there is no one else out there above you to help. Of course, there is a CEO, but her time is very expensive and limited, so her help should be more of an exception.

Applied tasks

The analytics director doesn’t deal too much with applied analytical tasks besides important research. Based on it, the top management will make important decisions.

What are the important issues in which the chief analyst is directly involved, but which are not part of building the system? Here are some examples:

• building a KPI system for the entire company and setting goals for them
• evaluating the profit from projects and ranking them as part of a regular planning
• building predictive models of the company and the market
• research that can change the company’s strategy
• assessment of potential M&A deals
• integration projects based on M&A results
• projects about data with big regulatory risks, for example, compliance with GDPR, SOX and other expensive abbreviations.

Processes

In large companies, the director of analytics (like any other director) cannot control every issue or task that comes her way. But at the same time, her job is to make sure that the results of analytics work, on the basis of which decisions are made in local teams, can be trustworthy, the analytics itself is actionable and delivers measurable business value.

The director achieves this by hiring the right people, developing and motivating them, building processes within the company (in a broad sense, not only analytics-wise), improving the analytical infrastructure, and defining teams’ focus and goals. In order to find and hire the best candidates from the market, the director has to deal with the HR-brand as well.

The analytics department is the nervous system of the company. Thanks to analytics, management receives the needed signals and makes decisions. The director’s task is to organize the work of a large group of people so that this function is performed and ensures the company’s growth.

Summing up

Since I first formalized and implemented grades in my team, I don’t think I can live without them. Grades have settled so well in my mind as a manager that now I think about all the tasks related to people in terms of grades, be it recruitment, development, financial motivation, assigning tasks and responsibilities, etc.

If you manage analysts, I hope you will find this framework useful—and, at some point you will be as excited about it as I am.

It takes a lot of practice to become comfortable with the grading system. As a manager, you need to grade at least a couple of dozen employees to understand the nuances and some borderline situations.

If you are an analyst, you can use this material as a guide to professional and personal growth. Feel free to send this essay to your manager and discuss where you stand regarding these grades and what you need to work on to get to the next level at your next meetup. This system allows such conversations to be meaningful and fruitful.