Going into 2019, data science and machine learning is all the hype. Organizations in every industry are starting to realize that the key to success is held in the ability to make organization-wide data-driven decisions. Organizations previously stored data for regulatory reasons or because it was a necessary evil of doing business. Today, organizations are starting to understand that data has value, and the true market disruptors are organizations that store and analyze huge, huge amounts of data. The Googles and the Facebooks of the world are storing every piece of data they can get their hands on, and then using that data via machine learning to become increasingly sophisticated in how they engage with their customers and optimize their internal operations.

Feeling the pressure to harness the value of machine learning, many organizations go out and hire a data scientist. As a data science consultant, I get to speak with many organizations in many different industries on their experiences, and I can tell you that this approach often results in failure. But why?

If you don’t currently have a data scientist on staff, how do you hire a data scientist? Without the necessary skills to evaluate the knowledge-base of a data scientist, many companies make the mistake of hiring an individual who may speak very confidently and may have some past related experience, but ultimately does not have the expertise to do the job they were hired to do.

The hype surrounding data science is new, which means that it’s attracting many folks into the industry. But those folks are often green and even if they speak confidently during the interview process, they likely have very little professional work experience as a data scientist. And while it certainly takes a bit of grit to become a data scientist in the first place, everyone could use some mentorship and guidance when they’re early in a new career, especially when navigating competing priorities in terms of what they still need to learn. There’s also lots of misinformation out there regarding data science and engineering, and without the mentorship of more seasoned staff, junior data scientists can sometimes find themselves believing incorrect information.

Because data science is all the rage, data scientists are expensive! According to Glassdoor, the average data scientist takes home an annual salary of $120,000 per year. This figure doesn’t include insurance, 401k matching, annual bonuses, etc. And while it’s highly possible to find a data scientist for less than $120,000 per year, you get what you pay for. Data scientists who have professional experience in many of the skillsets required to be the sole data scientist at a company are commanding high salaries in the Denver, CO area. And back to the first point—does your firm have the skills to identify a data scientist whose market value is $120,000 per year versus a data scientist whose market value is only $90,000 per year?

Do you have the internal expertise to understand what your data scientist should be doing? Every data scientist has walked out of an interview and thought, “This company has no idea what data science even is.” These companies often end up hiring someone who does not have the appropriate skills and assigns tasks that are not data science. These companies are often drastically overpaying for junior-level staff.

It should be noted that not every company that hires a junior data scientist finds themselves in this position. There are certainly good experiences to mirror the bad experiences shared by our clients. However, if you find yourself in any of the positions described above, I always recommend hiring a data science consulting firm, such as Datalere, to help execute your first successful data science project. This allows organizations to do the following:

    • For less than the $120,000 that you’d pay a full-time internal data scientist, you can harness the expertise of an entire company. Datalere’s two managing partners have been working in the data space for a collective 30 years and Datalere has folks on staff who have been working in data scripting languages and implementing data models and algorithms for the past decade.


    • You can spin up and spin down your data science spending on an as-needed basis. Not every company needs $120,000+ in data science solutions each year. Hiring a firm allows you to only spend what you need to spend—if you only have a certain budget for a solution, you can hire a firm, such as Datalere, within that budget and pocket the extra cash.


    • You’ll get to experience a crash course in how a successful data science project can be executed within your organization, with expertise from a firm that has successfully executed data science projects across multiple organizations and industries. Successfully executing a data science project in an organization that is new to data science is difficult and often results in failure. Datalere has had the privilege of assisting multiple organizations in multiple industries with successful data science projects, which means that we have a unique view of what works and what doesn’t.


    • Your existing staff will become familiar with experienced data scientists, including which specific skills data scientists can be expected to possess and which skills data scientists cannot be expected to possess. This can avoid the pitfall of overpaying for a more junior-level staff.


    • This first data science project can act as “proof-of-value” for your management, which may provide the push they need to authorize an internal data science department. Data science projects almost always result in a positive ROI, especially in cases where the project was implemented successfully, under budget, and on time. A good first-time experience is typically the push that management needs to pull the trigger on investing in an internal data science department.


    • You’ll get experience in delivering an end-to-end data science solution, from data acquisition, data pipelines, database modeling, to productionalizing and end-user visualizations.



  • At Datalere, we work with each of our clients collaboratively, placing a high emphasis on teaching, coaching, and providing explanations in plain English. Our clients should walk away feeling like they have a greater understanding of the data science process and how their own future data science projects could be successful, with or without the continued support of Datalere.If you’re struggling with any of these common pitfalls or feel like Datalere could help your organization turn around a failed data science project, feel free to reach out for a free consultation. We’d love to chat about what value we could add to your organization.