Textkernel’s semantic matching engine fuels one of the most successful projects for Randstad Germany

Randstad Germany decided to incorporate Textkernel’s semantic search- and matching engine within their automatic pre-matching project (APM). The project facilitated the simultaneous search throughout several applicant tracking systems, with the aim of matching candidates to suitable projects.

With an average of around 57,000 employees and 500 branches in 300 cities and sales of EUR 1.97 billion, the Randstad Group is the leading personnel service provider in Germany. Randstad offers comprehensive HR services to companies in various industries. In addition to staffing, the Randstad Group’s portfolio includes, among others, personnel recruitment, HR solutions and inhouse services.

Efficient temp work requires intelligent matching

At the present time temp work and staffing plays an important role. In addition to flexibility it also creates new jobs. In case of illnesses, higher workload or bottlenecks, companies such as Randstad provide suitable staff. More and more companies rely on temp work compared to time-consuming recruitment. In order to achieve this at the highest level, it is important to deliver suitable candidate profiles as fast as possible to the client. Therefore, matching plays a significant role.

Stephan Reiche, manager of marketing technologies at Randstad Germany, explains: “Automatic pre-matching (APM) is truly a milestone and one of the most successful projects in recent years. APM provides the basis for future digital processes and services and will definitely result in additional projects. With APM, Randstad is able to process the large applicant market more intensively, more precisely and in a timely manner.”

1,500 Randstad users are now using Search! and Match!

After an intensive project and some new Randstad-specific developments, the team has developed a completely new data router which provides data from various applicant tracking systems in Textkernel’s Search! and Match! user interface and thus offers Randstad the opportunity to search for suitable candidates simultaneously in their numerous systems. More than 1,500 Randstad users exploit Textkernel’s semantic technology to connect candidates and jobs faster. “The feedback from users is very positive and there will certainly be further developments. The speed ​​and efficiency of Textkernel’s semantic matching functionalities have inspired us, ” says Stephan Reiche.

“The feedback from users is very positive. The speed ​​and efficiency of Textkernel’s semantic matching functionalities have inspired us!”

The results

The results of the automatic pre-matching project are extremely gratifying. With Textkernel’s semantic search engine five systems can be queried simultaneously and semantically. One keyword entered by the user is linked to an average of ten synonyms, resulting in suggestions from various internal and external (e.g. LinkedIn) databases. For example, if a Randstad user enters the job title ‘carpenter’, Textkernel automatically supplements this with ‘woodworker’ and ‘joiner’, thus providing more results.

Above all, the speed is an important factor. A search query takes an average of less than 0.1 seconds, which means that incoming customer requests can be processed more quickly. “Matching processes are significantly accelerated. We also saw the first positive results in our sales activities. We can now match and recommend candidates with a one click to a suitable vacancy – and vice versa,” Stephan Reiche concludes his feedback.

“Matching processes are significantly accelerated. We can now match and recommend candidates with one click to a suitable vacancy – and vice versa!”