How can you coach less experienced data analysts to success?

Coaching less experienced data analysts involves a structured approach that begins with assessing their skills and areas for improvement. Clear, achievable development goals are set in alignment with their career aspirations and organizational needs. Access to mentors and training resources is provided to enhance their skills. Regular feedback and practical projects offer hands-on experience, while promoting continuous learning and soft skills development fosters growth. Regular check-ins, recognition of achievements, and long-term planning keep them motivated. Providing support and being patient are essential, as learning and growth take time. Effective coaching is a commitment to their success in data analysis.

According to me the success mantra is, work with the people on ground who has generated data and understand the need of audience, dont study data from 30000 Feet, be realistic and be on ground. And once you know what to read and what to deliver, take small steps and keep adding layer to layer to your output.

A great way to check analyst's strengths and weaknesses is using a simple checklist of the most important skills: 1. Product knowledge / domain expertise (no great analytics can be done if an analayst doesn't know the object of their research) 2. Analysis design 3. Data collection 4. Data analysis (descriptive statistics and statistical inference) 5. Data visualisation 6. Data storytelling 7. Programming (depending on your data stack: SQL, python, spark, etc.) It's also nice to know soft skills - often technical skills are there, but the analyst doesn't realise that their communication skills don't work. Try Gallup Strength Finder or, if your budget is limited, MBTI to point them to the right direction.

Determining the right level of coaching can be challenging. Should you guide someone closely from beginning to end, or should you give them more freedom? The ideal approach probably lies somewhere in between. I tend to give people the autonomy to discover and problem-solve on their own. I firmly believe that problem-solving and the ability to work independently are crucial skills to develop. However, it's equally important not to leave individuals feeling unsupported. They should have some direction and know that they can always reach out for assistance. Ultimately, it's a collaborative effort.

One thing I’ve found wildly helpful is running SWOT analysis on myself. Every day (or week if I’m too busy on a daily basis) I’ll assess strengths, weaknesses, opportunities, & threats… objectively. Here I gain clarity on my biases and can now assess SWOT on others understanding my own handicaps. Some things I find in others might be attractive to me because I lack in those areas… that doesn’t mean that the analyst in this use-case is best fit for this project. Bearing this in mind, you can help guide analysts in knowing what work they will be effective in. You can try to teach a fish how to climb a tree, ultimately this fish would likely experience imposter syndrome. We’re no different.

Allow them to fail forward, give them tasks, projects, or assessments, and provide feedback on what they did right, and wrong and where they can improve on.

Rather than providing immediate feedback, allow the individual to reflect on what they did well and where they could improve. Too often, leaders rush to a framework or model to provide feedback when we should elicit feedback directly from the source. After they deliver their feedback, you can provide insights and ask additional open-ended questions to have the person reflect further. You're trying to develop a self-serving team, not a group that always looks to you for judgment.

Giving feedback can be awkward and backfire - have you ever been in a situation where you unintentionally offended someone instead of helping them? To prevent it, it's best to set up the whole situation in as non-personal way as possible: describe the situation as a job to be done, and discuss how various actions influence the task positively or negatively. Focus on the object (the task) not the subject (the person).

Less experienced analysts don't know what they don't know. Pair them with a more experienced analyst to learn modeling and determine if they are stronger in code or visualizations. Then, give examples and feedback to help them grow.

As a coach, it is important to provide feedback but be very constructive about criticism. Make your feedback clear and give them areas of improvement. Also, follow up with them and make sure they implement your feedback successfully.

Live code reviews go a long way in helping junior team members learn best code practices. I use them often as a dual opportunity. For newer analysts, they learn from my code style and best practices. For me, I get fresh new perspectives on how my code is received and understood. It’s a great way to both make your work better and help junior analysts learn.

"Do as I say, not as I do." Everyone knows the old saying truly means the opposite and holds in business. Your team and those around you will learn and assimilate to the environment you create and enable.

Let's say you are developing a roadmap from your present level of data analytics?knowledge and abilities to more advanced data crunching skills. This roadmap consists of several pillars, and each pillar stands for a certain topic that needs to be mastered. The?bridge can fall apart if you attempt to build too many pillars at once or if your focus is diverted. Instead, concentrating on improving one pillar (or talent) at a time will guarantee a strong and effective bridge. Similarly, in the field of data analytics, it is wise to thoroughly understand one topic before moving on to the next. This method results in a deeper comprehension and?creates a solid base for future development.

In the early stages of my career, I learned the importance of code clarity and understandability when I was reading another person's code and I couldn't understand it. Code indentation, much like well-structured writing, aids in comprehending a program's logic. Comments serve as explanations, providing insights into the 'what,' 'why,' and 'how' of the code. These practices not only enhance code readability but also foster effective collaboration among team members. If I pass my code to someone, they should be able to understand the logic, and understand the structure and pick up where I left of right away! That is the ultimate goal of best practices.

I agree and they will learn faster if you let them do the hands on and provide a safe environment to do trial-and-error. Guide and show them some techniques on how to navigate complex problems, for example, ask them to write their thought process using a process flow chart. Then have a brainstorming discussion to challenge the process/logics. For coding, encourage them to start with the basics and as they become comfortable, slowly teach/guide them on how to write more complex codes. When they have strong foundations on the basics, it will be easier for them to do debugging as well as to build more complex/neater codes in the future.

Collaborating with other data analysts across your organization can provide a fantastic support network. Although some challenges may be domain-specific, many common challenges span domains. By exchanging ideas with other analysts, you may find ways to tackle challenges across the organization.

Promote a culture of continuous learning and self-improvement. Encourage them to explore new tools, technologies, and data analysis techniques on their own. Running a team in a pod structure helps to pair up analysts while giving them an experience of working with different stakeholders and expanding their knowledge base in multiple areas, while promoting collaboration.

Help them develop essential soft skills, especially communication and active listening, and coach them how to present their findings effectively to technical and non-technical stakeholders.

Gradually introduce more complex and challenging projects as they gain confidence and experience. This allows them to apply their skills and learn by doing.

There is no equal to hands on learning. Encourage independent attempts to complete tasks and offer a review with appropriate constructive feedback and praise.

In my experience, it is helpful if you apply all applicable models to the dataset. It is because many a times you get some invaluable insights from some of the models which you didn’t expect

Remember: there is no experimentation without psychological safety. Create an environment where errors and mistakes are cherished as learning points. Don't blame people for mistakes, instead treat them as challenges. Mistakes happen all the time, and the goal is to try to repeat old mistakes as little as possible, and get new mistakes. It's a bit like debugging a piece of code: be happy when you encounter a new error communicate, instead of blaming yourself for making an error.

In my team, we employ a simple yet effective coaching approach encapsulated by the acronym "BCD": B - Business Acumen: It's essential to grasp the intricacies of the business and industry at hand. Without this foundational understanding, one cannot formulate pertinent questions for analysis. C - Context Awareness: Dive into the context and background of the requirements. Craft a narrative that not only explains the 'what' but also delves into the 'so what.' D - Data Mastery: Just as a skilled chef knows their ingredients and their sources, a proficient data analyst must immerse themselves in data. A thorough understanding of data sources, lineage, and quality is indispensable for crafting insightful analyses.

We onboard around a dozen graduates every six months to our analytics division. Some of the things that have worked well for us include: - setting clear expectations: realistic goals provide a roadmap and a reference point to gauge progress - encourage curiosity: to foster a mindset of exploration we like to be quite hands off, and let them have the autonomy to dig deeper - peer support: in addition to having a buddy, we setup a teams channel so that anyone can ask any question, and we also have weekly team meetings where people are asked to present things they are working on - so that everyone can benefit from the experience - being nice: a fun working environment takes very little effort, and has great payback

Remember that when you’re coaching, the goal is not to have someone mirror your approach to data analysis. The goal is to make sure the people you are coaching are confident, empowered, and understand how to do the work properly.

New analysts have a wealth of specific knowledge and offer a great mentoring opportunity for you to learn together and forge a relationship. Often, a less experienced analyst has a wide range of skills they can apply but may need more context in using these skills outside of a lab environment. A more experienced analyst is invaluable here to guide the application of methods, models and tools and tell the data story. Developing the soft skills of a junior analyst is essential to their growth.