At Emsi our ambition is to bring our customers the best possible data set — the most robust, reliable, rich and relevant intelligence on local labour markets. From the beginning of March, our data will undergo its annual refresh, and this new data will be available to all our Analyst customers. But as well as this update, we’ve also been making some real in-roads on our ambitions in recent months, and we want to share with you how we’ve been moving the needle on our Labour Market Data and how we plan to continue this in the months ahead. Below are six of the most important improvements we’ve made.
Re-engineered Job Posting Analytics
We introduced JPA in 2017 and improved it gradually as we have grown our library of postings — we now have over 40 million unique postings, with vast reserves of details on roles, places, skills and companies. But we knew we could do better and at the end of 2019 we rebuilt our system from the ground up — some of the changes are subtle, but some are more striking.
On the subtle side, Emsi customers can now have much finer detail in their JPA queries: including and excluding different skills, or subsetting them around the number of days the postings were online. On the striking side, we’ve added new metrics around posting intensity — the ratio between unique postings and duplicate appearances — and this month we’ve added salary data, allowing you to see the median advertised salary and its spread, offering rich potential to understand how employers are positioning pay in local and skills-specific recruitment markets.
Improved Labour Market Information
While new ‘big data’ sources like Job Posting Analytics offer unparalleled richness, Emsi’s view is that traditional labour market information (LMI) based on robust survey data cannot be beaten for quantitative rigour: the real power is in combining LMI with the ‘big data’ sources, and for that reason we’re continuing to invest in updating and growing the data sets we build on those made available by ONS and others.
In the past year we’ve added new modelled sets around local area educational attainment and economic activity. While these measures are available from Nomis at local authority level, some of the smaller authorities see a lot of volatility and some missing cases, and we use modelling techniques to smooth out and make the data consistent.
Our new datarun is based on our modelling enhancements to the latest available release of the ONS’s Business Register and Employment Survey, collected in 2018, with our projections going out to 2028. Because we update our LMI data set each year, we make annual additions to our projections.
Introducing Online Profiles
Job Posting Analytics provide deep insights on employers’ demands for skills, but what about supply? To tackle that gap, this year we’re introducing Online Profiles data. Drawn from a variety of opt-in sources we de-duplicate across them and have in excess of 100 million profiles with a range of typical CV descriptive content: employer, role, location, skills, and educational background.
Already available in first form through an upgrade to Emsi’s Analyst tool, Profiles can be searched in similar fashion to job postings to find the pool of talent in a region and the kinds of skills and backgrounds they begin. Beneath the surface, the data offers further potential, with many profiles having further data on past jobs, allowing for Emsi to provide analysis on typical career paths and their intersection with different skills.
For some years, Emsi has been holding off on moving to more current local authority geographies in our data, because we value keeping our taxonomies stable where we can — but the time has come to change, and in the next month we’ll be moving to the 2015 local government boundaries, with a full dataset on 380 local authorities in Great Britain; an upside is that we’ll be moving over to the current GSS codes, making it easier to link Emsi data to other sources.
Far bigger news is that in that new release we’ll also be adding our range of intelligence for 11 completely new local authorities, in Northern Ireland. The absence of Northern Ireland from our data has long been something we wanted to tackle, and in 2020 it’s happening: Analyst users will have access to full UK-wide intelligence.
Higher Education Intelligence
Most of our customers have a close interest in education, and extending our data into education is always something we want to do. In 2020 we’ll be adding in rich intelligence on the volumes of subjects and qualifications taken across the UK Higher Education system, with data going back to 2013 and to the most recent available year. We’re already thinking about the reporting possibilities and how best to link that data to our wider range of labour market insights, but it will give a whole new dimension to Analyst, especially for our university customers.
The Emsi Skills Library
In 2019 Emsi made access to its skills library open for users in different environments across the world. The skills library has nearly 30,000 individual skills identified from a wealth of job postings and online profiles data. It’s the same library we use every day to identify skills sought by employers in job postings, and skills claimed by professionals in online profiles, and we update it every two weeks as customers spot skills we’ve missed or can better capture.
As well as making the library available, we’ve also allowed potential users limited free access to the API we use to try out our implementation — available at the link above. Into 2020 and beyond, we see skills as the critical dimension not only in understanding labour markets, but transforming their operation and we are already exploring the possibilities as to how we can bring the skills layer into our customer worlds across education, economic development and the workplace.
If you have any questions about our data and the improvements we are making, get in touch and we’ll be happy to help.