Looking to hire data scientists? Harvard-Incubated Experfy connects you to over 30k data experts for your data science projects. See how companies like Macy's, Vistaprint, Keurig and others use Experfy to connect with data science freelancers.
Data Scientist was considered the hottest and most in-demand job of 2017, and the trend will continue into 2018. Currently, there is a growing need for these specialists as more companies are looking to leverage analytics and data to get a competitive advantage. Decisions are no longer taken by gut feeling but are built on previous experience and looking at market trends. Artificial intelligence, machine learning, natural language processing, sentiment analysis and more are just a few of the techniques which are generically called Data Science.
The real question is: what is the best choice for your company regarding these services? Should you train your existing staff, hire data scientists or outsource to a professional organization? There is no single correct answer to these questions, and each entity should start with an evaluation of their expectations and needs. In this article we’ll provide some guidelines to facilitate this decision.
In-house or rent?
Once you have decided that you need a data scientist or a whole team of them it’s time to go on to the next step and decide whether to hire or rent. This is based on the scope of your current project, as well as further developments. Even for an in-house team you still need to decide between enhancing the skills of your current employees or bringing a new expert onboard.
If your employees already have some knowledge about data analysis, visualization and are accustomed to business intelligence products, you could consider sending them to a training session. Know-how in data science doesn’t come cheap, but it might be well worth it since you already have these people on your payroll, they know the business and are accustomed to working with each other.
You must select the most enthusiastic and well-motivated person or persons for this project and ensure they have a valid interest in sharpening their skills.
If your current employees don’t have the know-how or motivation to join a data science training program or you can’t afford to take them out of their ongoing work, hiring experts is another solution. However, be aware that qualified personnel in this area start at about six-figures, and usually you get what you pay for. Even if you have the budget, you might still have a hard time finding the right people since there is a small pool of experts available and a high market demand.
Having your in-house data expert could be a starting point to grow a whole team. The sheer complexity of using data science calls for a dedicated team and just preparing the information for analysis takes up a considerable amount of time.
A more convenient solution if you just want to test the benefits data science can bring to your organization is to hire a dedicated company and outsource the entire project. It is not the cheapest solution, but it offers the guarantee of quality, reliability, and access to a team with a wide range of skills.
A successful project usually requires at least six to seven different skills (machine learning, analytical approaches, management, engineering, data governance, visualization and domain knowledge). It is difficult to find that Data Science Unicorn, a mythical creature that can perform the analysis and interpret the results by using their excellent communication skills.
A more down to earth approach suggests dividing the work between five to six team members and combining their abilities to create synergies.
Fear of missing out
The FOMO concept comes from the investing world but applies here as well. Companies could rush into creating ad-hoc teams to get some data analytics capacity, without proper training or planning. Hiring a company which provides these services is not a better choice if your organization has not decided the scope, budget and realistic goals of the data science project.
Regardless of the choice you make, hiring or entering a partnership with a data science company, don’t just act on impulse. Create a plan outlining the pros and cons, requirements, expected results, necessary skills and possible drawbacks. Consider both the short-term and long-term budget as well as your overall strategy, not just your current project.
The right choice
Before making your decision, look at the data. Describe the type of results you expect. You could even go to job sites and get some inspiration regarding the requirements and tasks you expect your data scientists to be able to perform. By going through this list and your current team, you can decide if you want to go for the training approach.
If you are looking to scale up and become an industry leader, most likely you will not have the time to take it the slow way and create data scientists. In this case, it makes much more economic sense to rent the analysis capacity of a dedicated company which already has in place the specialists and processes.