So many great points are made in this article by HBR.org for why data science teams should consist of more generalists than specialists, which I think can be applied to other analyst groups across your business.
The division of labor Adam Smith demonstrates in The Wealth of Nations is so ingrained in our society today, that many businesses look to organize their teams in the same fashion. Data science teams shouldn’t be optimized for productivity gains; the goal is not to execute. The goal is to learn and develop new business capabilities.
According to the article, specialization hinders your team’s goals in three ways:
1. It increases coordination costs.
2. It exacerbates wait time.
3. It narrows context.
Because your team consists of a bunch of specialists, it may take a lot of time to organize meetings, discussions, or reviews. It may be hard to coordinate each efforts just because each specialist may be allocated to several other projects. If you had a data scientist with more general knowledge, it may be more efficient for them to work alone or coordinate with one or two more analysts, rather than having a bunch of specialists in the room.
I think the points made in the article can apply to any group of analysts. For example, a financial analyst may be really efficient in using excel, but what if you do this analysis frequently, and it may take you hours, even days just to do it? It would definitely be useful to have a financial analyst who at least has some understanding of creating macros or writing VBA code to automate their process. If they don’t have this general knowledge, they waste their own time and their companies time doing the same repetitive tasks over and over again when it can be done more efficiently.
Even if they decide to improve on this process, they will need to spend more time finding the right person or people in the company who have this knowledge and coordinate with them to discuss the issue at hand. Even after finding the right people to work on this automation process, it may take longer than expected to have a finished product at hand because the people who is helping the financial analyst will usually have other obligations to attend to first. Therefore, instead of the financial analyst having knowledge about Excel macros and getting this process done in a couple days, they will have to wait weeks or even months because they need to rely on others.
The article closes with more great reasons on downsides to having specialists. The author makes the point that having specialists can lead to a loss of accountability and passion from workers. The division of labor leads to the dulling of talent, workers become ignorant as their roles are confined to a few repetitive tasks. Specialization may provide process efficiencies, but it is less likely to inspire workers. On the other side of the spectrum, generalists roles provide things that drive job satisfaction: autonomy, mastery, and purpose.
I encourage you to read the actual article, as it not only explains why the division of labor does not work in today’s work environment, but it explains why having people with a breadth of knowledge will also help with job satisfaction, and having people passionate about what they do, will make a huge impact on your business in the long run.
Source: Why Data Science Teams Need Generalists, Not Specialists
Commentary: Why Data Science Teams Need Generalists, Not Specialists
So many great points are made in this article by HBR.org for why data science teams should consist of more generalists than specialists, which I think can be applied to other analyst groups across your business.
The division of labor Adam Smith demonstrates in The Wealth of Nations is so ingrained in our society today, that many businesses look to organize their teams in the same fashion. Data science teams shouldn’t be optimized for productivity gains; the goal is not to execute. The goal is to learn and develop new business capabilities.
According to the article, specialization hinders your team’s goals in three ways:
1. It increases coordination costs.
2. It exacerbates wait time.
3. It narrows context.
Because your team consists of a bunch of specialists, it may take a lot of time to organize meetings, discussions, or reviews. It may be hard to coordinate each efforts just because each specialist may be allocated to several other projects. If you had a data scientist with more general knowledge, it may be more efficient for them to work alone or coordinate with one or two more analysts, rather than having a bunch of specialists in the room.
I think the points made in the article can apply to any group of analysts. For example, a financial analyst may be really efficient in using excel, but what if you do this analysis frequently, and it may take you hours, even days just to do it? It would definitely be useful to have a financial analyst who at least has some understanding of creating macros or writing VBA code to automate their process. If they don’t have this general knowledge, they waste their own time and their companies time doing the same repetitive tasks over and over again when it can be done more efficiently.
Even if they decide to improve on this process, they will need to spend more time finding the right person or people in the company who have this knowledge and coordinate with them to discuss the issue at hand. Even after finding the right people to work on this automation process, it may take longer than expected to have a finished product at hand because the people who is helping the financial analyst will usually have other obligations to attend to first. Therefore, instead of the financial analyst having knowledge about Excel macros and getting this process done in a couple days, they will have to wait weeks or even months because they need to rely on others.
The article closes with more great reasons on downsides to having specialists. The author makes the point that having specialists can lead to a loss of accountability and passion from workers. The division of labor leads to the dulling of talent, workers become ignorant as their roles are confined to a few repetitive tasks. Specialization may provide process efficiencies, but it is less likely to inspire workers. On the other side of the spectrum, generalists roles provide things that drive job satisfaction: autonomy, mastery, and purpose.
I encourage you to read the actual article, as it not only explains why the division of labor does not work in today’s work environment, but it explains why having people with a breadth of knowledge will also help with job satisfaction, and having people passionate about what they do, will make a huge impact on your business in the long run.
Source: Why Data Science Teams Need Generalists, Not Specialists
Commentary