Are You the Only Data Scientist in a Small Organization? Here Are 5 Tips to Put You on the Path to Success
Originally published here
Be resourceful, stakeholder management, and continuous learning
Are you a data scientist working for a small organization? It’s an interesting position to be in. It is ideal for those who enjoy analyzing data and business problem-solving, as a data scientist will be expected to do both of these tasks on a daily basis. Often, you are running a one-man show. Executives know the value of having an analytics capability, which is often not their strength. The task is given to you to solve most data-related business problems.
As a data scientist for a small organization, you are expected to work across many parts of the business, including stakeholder management. There are usually no dedicated data engineers, data security personnel, or project/program managers in these organizations, so you will need to wear many hats when necessary. This can be a challenge, but it is also an exciting position to be in, as you will have a lot of autonomy and will be able to see the impact of your work firsthand. You are not given large budgets for either time or tools, but being scrappy is part of what makes this job interesting.
In my 20+ year career, I have held many formal and informal titles — data guy, analyst, IT guy, developer, statistician, and data scientist. I am sure many of you can relate. For this article, I want to focus on an equally important role — the “only” (or lonely) data guru or whatever you like to call it person in a small organization. Organizations of all shapes and sizes are finding themselves wanting to take advantage of the spoils of big data analytics but not having the resources — time, budget, personnel, skillsets — to be able to do it themselves. Here are 5 tips that you should consider to succeed in this environment.
1. Establish yourself as the data expert
Is this too obvious? Perhaps. Often we have chosen this profession because of our love for data and developing something useful to solve problems. This also means that we spend hours and days in front of our laptops and assume that others will have some understanding of all the complex hurdles that we had to overcome to produce reports, insights, or applications. There is no established data science team(s). It’s just you. To be considered a valued member of the organization, you have to be respected for your expertise.
That’s easier said than done. What has always worked for me is getting a list of the most pressing problems that the executives wanted me to solve. Find a relatively small but high-impact problem. Nail it and make sure that it is recognized as solving a high-priority problem.
I also said “small” because you must produce something quick initially. This can’t take months. Based on your first project, the impression should be that you can solve critical problems quickly because of your expertise in the data area.
When people see you as an expert, your visibility with management will increase, and they will start to seek out your expertise even more. This will lead to getting the most critical stakeholder’s support — the executive support.
2. Use resources available to you
You have a very limited budget for your work. You will have to be frugal and show that you can deliver by utilizing what’s readily and freely available. Fortunately, we have access to a wide range of open-source tools such as python. You can set up a robust ETL process and automate data cleaning standardization and visualization using python. Also, don’t be afraid to present the results in Excel as most of our stakeholders are comfortable with it anyway.
Demonstrating frugality and resourcefulness in the past may assist you in obtaining more resources in the future. Small organizations often do not have a dedicated IT/data project budget. They are all included in the operational budget so that you will face resource constraints from the beginning. Creating this perception that you are frugal can add more weight to additional resource requests in the future.
3. Find out who else has an interest in analytics
I have always found that this is a critical ingredient in having my work being accepted and being used throughout the organization. Instead of advocating for the value of my work, I develop a strong relationship with those who are interested in analytics first. They will become your champions. Your champions will promote your work and create support for current and future projects. This will address one of the most critical aspects of succeeding as a data scientist — Stakeholder management.
3 Critical Aspects of Effective Stakeholder Management For Data Science Projects
4. Develop a strong working relationship with your IT team
Most of your work will be done on a platform that your IT team is managing. When facing resource and time constraints, you will need a strong partner who can take away obstacles for you. They might be able to open up ports so you can interact with external systems. They might even let you remote access one of their servers so you can run resource-intensive analytics processes. What I see as the most important benefit of working closely with the IT team is their timely response to my cry for help. I prefer not to make a formal request and wait a week or two to get something done. Their willingness to help and get things done efficiently and quickly can advance what you are trying to achieve.
5. Network with other data scientists
This is a crucial step that might be the difference between success and failure for you in the long run. You don’t want to go through this entire experience alone. It’s in our nature to find our groove and be content with being a highly regarded member of an organization. In other words, it’s easy to be happy with what we have and lose the drive and urgency to develop new skills. One of the downsides of working in a small organization is the lack of diversity in the type of problems that we are trying to solve. Look externally to see the current developments in the data science field. Frequently check Kaggle to see what others are doing. Join a Facebook group to interact with other data scientists.
I have volunteered to present at a virtual workshop which led me to several conversations with data scientists in other industries. The learning has to continue, and we will most likely have to look externally.
A Hiring Manager Who Went Through 120 Resumes for a Junior Data Scientist Role
There are a lot of resources available in how to succeed as a data scientist, but most of the advice is geared towards larger projects/organizations with larger budgets. The five tips I’ve outlined should help put you on the path to success if you are working as a data scientist in a small organization. Develop a strong working relationship with your champions and IT team so they can remove any obstacles for you and allow you to work more efficiently. Network with other data scientists so that you can learn from their experiences and stay up-to-date on the latest developments in the field. And continue learning; never stop developing your skillset, as this is what will set you apart. With these tools at your disposal, you’re sure to make great strides in your data science career, even if you are not working for Google or Facebook.
Comments ()