Demand for professionals in data science and analytics is expected to rise significantly over the next years (cf. this study by IBM). In order to keep track of future job trends, we started the DataCareer Job Market Index (DJMI) in July 2017. We track job openings on the biggest online job board, Indeed, in the fields of data science and analytics, data engineering, business intelligence, artificial intelligence and statistics. Sign up for our newsletter if you’d like to receive a monthly summary of the results.
We searched for jobs on Indeed using the following search terms: Data Scientist, Data Analyst, Data Engineer, Big Data, Machine Learning, Artificial Intelligence, Business Analyst, Business Intelligence, Statistics, Quantitative. Using the jobbR package, we downloaded the jobs into R for further processing. Since the search returns some results we'd like to omit from the job index (e.g., general Scientist or Software Engineer jobs), we filtered these jobs out. The index only counts jobs that contain at least one of the following expressions in the job title: Data, Analy, Business Intelligence, Machine Learning, Deep Learning, Artificial Intelligence, AI, Computer Vision, Neural Net, Natural Language Processing, NLP, Hadoop, SQL, Oracle, Statisti, Quant, Bioinform. This filters almost all jobs correctly, although it will miss a few data jobs that lack these keywords in the title (e.g., some general PhD student ads).
The search is performed at the end of each month, with only jobs from the last 30 days being included. Using this method, how accurate can our job market estimates be expected to be? While we are likely to underestimate the absolute number of jobs, the relative changes we record over time should provide a very accurate measure of job trends. We therefore indicate relative changes compared to the reference point of July 2017, which is set to 100 in our market index.
In Germany, job ads increased by 36% between July 2017 and January 2018, with the biggest rise occurring in January (+17%). Interestingly, we did not observe a drop in job openings during December. In July 2017 we counted a total of 2’492 jobs.
In Switzerland, data job ads increased by 28% between July 2017 and January 2018. The drop in December (-13%) can likely be explained as a seasonal fluctuation, given that companies publish less job openings during the holiday season. During our reference month, July 2017, the index included 544 jobs.
In the biggest data job market, the US, jobs remained relatively stable between July and December 2017. In January 2018 we observed a 12% increase in jobs. The total number of US data jobs was 35’947 in July 2017.
In the UK, jobs increased by 24% in January 2018 after a drop of 18% in December 2017, . Again, we expect this to be an effect of the holiday season (which apparently leads to either a significant decline or stagnation). In July 2017, we counted 7’633 jobs for the UK.
If we consider the job counts for January 2018 and relate them to population size, we get the following number of jobs per million inhabitants: 125 for the US, 125 for the UK, 81 for Switzerland, and 41 for Germany. However, these numbers should be taken with caution, as the percentage of jobs that are published on Indeed may differ between the countries. We’re currently conducting research into open questions such as the popularity of Indeed in each country and the factor by which Indeed-based estimates underestimate the total number of jobs.
For questions, comments or if you’d like to collaborate on our research, don’t hesitate to send us an Email at firstname.lastname@example.org.