Using data to improve job placement: optimising job hunts through research and platform design

Digital technologies have fundamentally changed how jobseekers and employers find each other in the labour market. Online job platforms are a key element of this matching process.

This research project, led by Michael Siegenthaler (ETH Zurich), examined for the first time how jobseekers in Switzerland search for jobs online – with the aim of better understanding job searches and improving job placement.

The centrepiece of the project was novel click data that shows how jobseekers navigate job listings, what they look for and what they click on. This data, combined with information on job listings and social insurance register data, provides new insights into search behaviour on the Swiss labour market.

One of the key project outcomes was the development of an interactive dashboard – www.swissjobtracker.ch – an instrument for the real-time observation of Swiss labour market trends.

The most important findings

The research findings show, for example, that the search radius has a major influence on finding a job. Those who cast their net wide benefit from a larger selection of potential jobs. However, this does not necessarily mean that expanding the search radius is beneficial to all jobseekers. Those who search in a targeted way are often more efficient and can more quickly convert newly created jobs in their field into actual employment. The search radius varies by jobseekers’ sex, level of education, origin and family situation. For example, women with childcare responsibilities attach greater value to a short commute. These patterns suggest that a successful search radius is very individual. They also help to explain why labour market policy measures that seek to expand the search radius without taking account of individual needs are not always successful.

Fringe benefits are sometimes more important to jobseekers than the salary

The click data also shows that jobseekers’ interest in a role is not only influenced by the salary, but also by fringe benefits such as the opportunity to work from home, a company car and childcare. In a broad-based online field experiment, the researchers were able to prove that jobseekers are willing to accept a salary around 15% lower if in return they are able to work from home or get a company car.

Significance for policy and practice

The research project clearly shows that the most important thing is not only how many vacancies are advertised, but whether jobseekers are even aware of them. This gives rise to concrete courses of action for policymakers, public employment services and employers. The studies demonstrate how data-based insights from online job platforms can be used to provide more targeted support for jobseekers, to make platforms more user-friendly and to identify labour shortages at an early stage.

Three main messages

1. The research findings show that the design of job platforms – for example, default settings, search filters, the occupation classifications used, the way job details and recommended job listings are displayed – has a significant influence on which vacancies jobseekers see and which ones they apply for. Systematic scientific evaluations of the platform designs reveal significant and largely untapped potential to make the matching of candidates to jobs more efficient.
2. Employers in Switzerland appear to hold significant market power. This is partly because jobseekers only rarely change occupation or region, have widely differing expectations of jobs, and are very unlikely to react to a difference in pay between two posts. Policymakers should therefore consider measures to increase competition for workers – particularly in low-paid jobs. For example, jobseekers could receive more targeted support to change occupation or region, and companies could be required to disclose pay in job advertisements.
3. If public employment services and researchers systematically collect, compile and analyse data on clicks, applications and job offers, this allows more effective monitoring of labour market trends and mismatches in the labour market. The weekly time series on www.swissjobtracker.ch illustrate this potential. The recording and analysis of this information also form the basis of in-depth studies on the characteristics, causes and consequences of job search behaviour.

To find out more about the exact methodology used by the researchers, and for more background information on the research project, visit the NRP 77 project website:

You can find other research projects on the digital transformation that are being conducted as part of National Research Programme NRP 77 here: