Section 1.1: Understanding the Challenges of Consulting

One of the most compelling reasons that led to my departure was the difficulty in perceiving the long-term impact of my work as a data scientist. This is a common sentiment among consultants, and I initially thought it was merely a convenient excuse for leaving. However, I came to realize it was a significant factor in my decision.

The nature of consultancy itself contributes to this feeling. As consultants, we provide expertise and recommendations to clients, but once the project ends, the solutions—often a model or analysis—are handed over for the client to execute and maintain. For instance, if I developed a demand-forecasting model, I could evaluate its performance using historical data before delivery. If the partner convinces the client to extend the project, I might receive additional data for assessment. Yet, I likely would not know how the model performed after its deployment, let alone its long-term effects.

The ongoing debate about whether implementation should fall under consulting firms' responsibilities persists. While some argue that consultants should focus on proposing solutions, others believe that without implementation, the efforts are futile. During my time, McKinsey increasingly emphasized implementation; however, those entering this field should not expect to see their models in production or measure their long-term success.

Section 1.2: Limited Exposure to Advanced Data Science Techniques

Consulting firms are typically engaged because clients lack in-house expertise in specific areas. As a result, many clients who seek analytical assistance are not at the forefront of data science innovation. Consequently, as a data science consultant, I found myself rarely working on advanced projects like recommendation systems for major tech companies.

While there are occasional opportunities to work on projects that utilize advanced methodologies, such as dynamic pricing, the majority of assignments often require simpler approaches. Moreover, clients just starting their analytical journeys tend to be skeptical of complex models, favoring transparency instead. This means consultants often have the chance to hone their skills in explaining analytical concepts to non-experts, which is invaluable, but may miss out on applying the latest advancements in data science.

This aspect may be more significant for some than others. If your passion lies in machine learning, AI, or advanced analytics, it's essential to consider this before pursuing a consulting career. On the other hand, if you're interested in leveraging data to solve tangible business challenges, consulting can be an excellent platform to develop those skills.

Section 2.1: The Pressure of Short-term Engagements

Due to the short-lived nature of consulting projects, clients often expect quick results, leading to a focus on immediate deliverables. This environment prioritizes efficiency over scalability, resulting in many analytics outputs being treated as prototypes rather than fully developed solutions.

Through my experiences in the tech sector, particularly in management, I recognized the significance of long-term projects. Such initiatives are typically spearheaded by leaders who proactively address broader company objectives. Conversely, consultants are usually reactive, solving specific problems without the opportunity to contribute to long-term strategies. This can be frustrating, as it limits the ability to understand and influence the larger context of the work.

Section 2.2: Personal Strain of Consulting

The final factor in my decision to leave was the impact on my personal life. Consulting is infamous for its demanding hours and lack of work-life balance. Burning the midnight oil to produce analyses and presentations is commonplace.

Additionally, the pre-pandemic norm required consultants to travel extensively from Monday to Thursday for client engagements. While the prospect of visiting different cities may seem appealing initially, the thrill wears off quickly when most evenings are spent working late in generic business hotels, leaving little time for exploration.

Summary:

As I mentioned in a previous article about my decision to join McKinsey as a data scientist, consulting offers a valuable starting point for aspiring data scientists, but it might not suit everyone. Key takeaways for prospective data science consultants include:

• Consulting may be more suitable for those who prefer exploration over building, as most projects involve handing off prototypes without further involvement.
• You will likely develop stronger data storytelling skills than advanced modeling techniques.
• Expect opportunities for efficient problem-solving on short-term projects, but fewer chances to engage in long-term analytics planning.

For additional insights into data science careers, consider exploring these articles:

• Why I Joined McKinsey as a Data Scientist
• Should You Ever Be a Data Science Consultant Despite the 80-Hour Work Weeks?
• 5 Lessons McKinsey Taught Me That Will Make You a Better Data Scientist
• 5 Mistakes I Wish I Had Avoided in My Data Science Career