Jacqueline Paige is the Managing Partner of Executive Recruiters, Smith Hanley Associates, specializing in the contingency placement of data scientists and other analytical specialists.
We recently checked in with Jacqueline to get her insight on the job market for data scientists and what those in the field should be doing to grow their skill sets. Here’s what she shared:
Can you tell us about the mission behind Smith Hanley Associates? What’s your approach to recruiting?
We’ve dedicated ourselves to being niche specialists. Since opening our doors in 1980, we’ve concentrated on just a few specific functional areas. We know these areas inside and out. We are focused on finding the perfect fit between clients and candidates for highly specialized positions within those highly specialized functions. We strive to make the recruiting process seamless.
What trends are you observing in the job market for data science and analytics? What do you predict for the field in the next five to 10 years?
Since we’ve been recruiting and placing statisticians since 1980 we’ve seen a number of trends come and go but nothing like the explosion of data from the internet and the impact that is having on the statistical career path. The demand for candidates is through the roof and only recently have academic programs started to answer the need to train math and stat experts for true data science work. Here are a few of the seismic shifts:
1. There is so much data the analytical expert can no longer pass off the management of that data to the IT department. To be a data scientist, to work with big data, candidates must have advanced IT skills beyond SAS or R. Hadoop, Python, Pig etc. must be in their skill set.
2. Becoming the top analytical person or part of the top analytical group in any company no longer requires a Masters in Statistics. BS or BA in a quantitative area plus specialized training that is being offered in 12-week courses for data science can be sufficient to take the next step.
3. For business and marketing applications the more personable, savvy and articulate candidates (with great stat skills) were often the most in demand. Now and in the future candidates that can manage and interpret the data through their IT skills (with great stat skills) will be the winners in the job filling game.
4. We are starting to see some resistance to candidates who need sponsorship. As more U.S. Citizens pursue this career path options for H1B or F1 visa holders becomes more limited.
5. In the latter part of 2016 the actuarial professional society began offering new testing and new credentials in predictive analytics and data science. Over 60 percent of P&C insurance companies consider themselves data driven, but lack of staff with the right skills is hindering their success. This designation will allow employers to quickly identify those professionals with the skills they are looking for and encourage more actuaries to get these skills.
In five to 10 years the academic programs and computer software will have caught up to the demand today and more well-trained candidates will be available … but today it is definitely a seller’s market.
What should data scientists be doing today to make themselves more marketable in the future?
Find a way to continually improve, expand, increase your software and data management skills. Commit to taking regular, formal training. Join discussion groups on the internet to stay abreast of what is the latest and the hottest. Career advancement is not going to be through a hierarchy promotion but through an in-demand skill set.
What advice can you offer data scientists on negotiating salaries and/or raises? What are the dos and don’ts?
Do: Work with a data science recruiter. They will know what is competitive and what isn’t.
Do: Don’t worry about title but do worry about who you report to. Some companies still have a fear of what they don’t understand, so make sure your boss is a champion of your area of work.
Do: Be reasonable but demand a competitive salary. In this market you can.
Don’t: Tell what your brother’s best friend’s sister makes as a data scientist. You must have more information than one anecdotal story
What are the most common mistakes you observe job seekers making?
1. Focusing too much on the salary and not on the viability of the company’s products and their role in making those products successful. If the company is successful, you will reap the rewards.
2. Assuming their skill set is top-notch. Too many great statistical candidates are missing the software and data management skills required for a true data scientist position.
What tools or resources should data scientists use to determine their value to a particular company?
1. Fit versus the job description
2. Input from a specialty recruiter.
3. Experience working at a competitor or in the same product area.
4. Years of experience in particular skill areas … not just years of experience in the job market. For instance, two years of machine learning or Hadoop experience is infinitely better than one year.
5. Cultural fit in the company. If you know people there, do you like them? Could you see yourself working there? Is the boss someone you admire? Do they support continuing education? Conferences?
What skill sets should data scientists consider harnessing in order to make themselves more valuable to their employers?
Flexibility and ability to learn new things quickly. This market is still changing dramatically year to year. R is hot right now but that could change in a few months. Hadoop, Python, Pig are so important, but chances are those names will be different in 2018. Finding a way to keep up with these new products and new applications every year, not just at the beginning of your career, is critical to long-term success.
Machine learning is so important and in-demand right now. Artificial Intelligence is going somewhere just not clear where yet. IoT is emerging as a significant opportunity due to the ease with which data can be transferred across networks. Insurance companies, agricultural equipment manufacturers and durable goods producers are using sensor data to predict part failures an aid in inventory planning and much more.
What would you love to see more data scientists doing to improve they’re ability to find new positions?
Use the latest and greatest applications in your CURRENT job. On-the-job experience is so much more credible and marketable than academic training. Introduce these products and methods yourself, if you can’t convince your boss or others in your group to pursue. If you have to get this experience through academic resources, get a consulting assignment in the real-world to apply what you’ve learned BEFORE looking for a new job.