Are you looking for people with Data Scientist skills to join your teams? Find out how you probably now have those people without noticing it.

In today’s world, data has become the best resource for every company. We all realized that, by using our own information and records, it is possible to design better plans, marketing campaigns, and forecasts, as well as generate revenue while adding value to the business.

In 2020, many companies started looking for this “new” role outside their organizations: the Data Scientist. These skills are now in big demand but difficult to find. Are the people in your company ready to take control?

Let’s look at the main requirements and Data scientist skills to consider:

Most common ones:

• Statistics & probability
• Data manipulation & analysis
• Data visualization
• Programming knowledge

Other skills required:

• Software engineering
• Model deployment
• Big Data
• Storytelling

We all agree that the first list is basic for a person to be considered a Data Scientist. However, if you take a quick look at the job posts on LinkedIn, you will see that many companies add the second list to the role, ignoring the fact that the person may need to know how to use about 50 (out of ~1,500) additional applications to demonstrate proficiency in those skills.

The problem is not the list itself, because companies really need the skills to handle tons of data that come in daily. However, after 10 years of experience working with data, I’ve noticed that people outside of the IT department but within the same company may have the first list of skills, but they’ve spent 80% of the time organizing raw data (according to the IDC, 17% of the time doing recurrent tasks, which is equivalent to one day a week). Bear in mind there is also another list of tasks that your development team has to handle:

• Spreadsheets
• Flat files
• Presentation slides
• SQL
• Data access restrictions

One important aspect to think of is that when you decide to hire people to work in your business, they likely have at least three skills from the first list, and as time goes by, they acquire the skills from the third list. Now, they know your company, share your vision, and probably swim within your organization’s processes easily (i.e., they make things happen).

Hiring an army of Data Scientists with the skills from the first and second list will put you up to speed with new ways to handle your data, but over time, the third list will overwhelm that new army. Daily tasks will consume their time and skills.

Isn’t that what is already happening?

Don’t miss this reading: How to become a Data-driven organization

Divide and conquer

I am not saying you are looking for the wrong skills; you really need to add them to your team, but the best team is the one where everyone does their part.

Do you have a person working as a Data Business Analyst (DBA), another professional as an expert on spreadsheets, maybe one SQL developer, or an economist? If you have any of these roles in your company, you are already on the right track, and the next step should be to remove the tasks that they were not hired for.

If you have statisticians, economists, or marketing experts on your team, they already have the knowledge the company needs to create models, forecasts, and insights with business-related value added. However, as I mentioned before, they spend 80% of their time acquiring, organizing, and storing data.

They also work locally in their workstations, which probably means that you might lose some data when anyone leaves the company. Have you ever had to save the hard drive files of the person leaving the company anywhere else? Did you lose that important reporting query? Every time a person within your company creates a report, they are using your data and creating new records; therefore, it is important to understand how data is created or transferred within your business.

The following is a very common way in which the data flows within your company:

• Acquisition
• Enrichment
• Transformation
• Reporting

If you see that any of these steps take place in spreadsheets in your organization, that is great. It means your team is doing enrichment, equations, or models that only need to be automated.

Now, by wanting to hire someone that fulfill the Data scientist skills, you may be thinking about these:

• Automated data enrichment
• Automated reports
• More accurate forecasts
• Machine Learning

As you can see, automation is a skill that you can easily find in Data Engineers, Data Analysts, and ETL Developers. They are professionals that handle tons of data, enrich, and move records from outside your organization to any data storage within it. Leading data companies such as Cloudera and AWS recommend these three roles for handling data automation, which is an addition (not replacement) to Data Scientists.

Generally, Data Scientists are more used to working in forecasts, deep insights, and (if it’s possible) Machine Learning algorithms. Cleaning the data is part of their work but is not what you probably would like to hire them for.

It seems that we need more than Data Scientists to automate data processing and fulfill one of the most important premises in business — reduce Time to Market. With Data Scientists, you will have better information, but it is not guaranteed they will make it on time. Let’s be honest, in today’s world, we need to make things happen fast.

Fight for freedom

At this point, we know we need different roles and people handling automation. Now, we can take that 80% of tedious data tasks away from our company experts and turn them into fully qualified Data Scientists.

Wait, didn’t I say they were already Data Scientists?

No, your company experts probably have at least three of the skills required, but it is also very likely that they miss the “programming knowledge” part. Therefore, if you want to automate data flows, you need to give your people the tools for accessing those pipelines. Believe me, it is much easier to learn to program for consuming data (only for consuming data) than to work with statistics, marketing, campaigns, or economics.

Adding Data Engineers, Data Analysts, and ETL Developers to the development team will take a big part of that 80% of tedious tasks away from your business experts, giving them fresh time so they can use it to learn and create. Let them work on what they need to learn: Python? R? Java? That’s up to them. Furthermore, allow them to ask new questions and not only work on what others are requesting. They will have new, organized, and cleaned data flowing into their hands, which means new insights, new enrichments, and even new products.

Your business experts can now look into the future because they won’t spend time organizing data. Your experts (already in the company) have more time to watch the current context and plan the future. Give them the power of free-thinking, give them moonshot goals, and challenge them to reach Mars.

This is your perfect data expert army

I’ve been saying many times that you need multiple roles, and you probably had in mind to hire ten new team members, but at this point, you probably agree with me. So, how should the team look?

Of course, it all depends on how you view your business and how much data you handle or plan to handle. Regardless of what you have in mind, you need Data Engineers and ETL Developers to automate your data flows. An ideal and typical team implementation (based on the 10 people you had in mind) will look like this:

• 2 Data Engineers
• 2 ETL Developers
• 4 Data Analysts
• 2 Data Stewards (You want quality and governance, don’t you?)

Do not forget that to do it correctly, you will need a methodology to sync the work of these new team members. Could it be Scrum? Kanban? I will address that in another post.

I hope the lines above help you build the team of your dreams!

Comments?  Contact us for more information. We’ll quickly get back to you with the information you need.