How much efficiency can we improve for an MNC?

Every year, Human Resource department of L’Oreal receives tens of thousands of applications from university graduates across whole China, all of them vying for limited number of positions offered by the company’s Management Trainee Program.

Colleagues responsible for campus talent acquisition used to handpick resumes from mountains of paperwork, spending a lot of time reading self-selling overstatement, which would later prove to have zero correlation with job performance.

Even though more than 40 people in HR department put in full energy to do the work, the amount is just too large. To make it a bit easier, they would narrow the search only to top tier universities in China, excluding, sadly, lots of potential talents who do not go to these universities.

Now the question comes: How can the HR department increase the talent-searching efficiency?


From year 2014, L’Oreal decided to do away with the tedious work of CV searching, and replace it with competency matching technology. Recruiters in L’Oreal co-worked with experts from Seedlink to generate three open-ended questions they would like the applicants to answer. These questions were tailored to discover a candidate’s communication skills, ability for team work, fit to the company culture, and sense of fashion.

After creating the position in RCXUE system, HRs published the position L’Oreal’s WeChat public account. All those who read the position info can apply for it through mobile devices. The answers from applicants were fed into the backend system of Seedlink, and the algorithm would analyze the answers and provide a ranking of candidates. Recruiters from L’Oreal can log into RCXUE platform to check the ranking and decide on who shall be given an interview.

In 2015, Seedlink helped L’Oreal add a new function to the platform – video question. It required a candidate to record a 30-second video to answer the question in English. By doing this, recruiters were able to gain a quick impression of the applicants and know their English-speaking proficiency.


Using the completely automated system, L’Oreal was able to cut its talent-searching time more than 10 times. Liberated from the time-consuming CV reading and filtering, recruiters now can spend more time on other parts of hiring process such as Skype interview and face-to-face interview.

On the other hand, in order to get a ranking, the system only required answers to the three open-ended questions, without taking into account other information like educational background. It means no matter which school an applicant came from, he/she could apply for the position. This helped L’Oreal to broaden its searching pool without adding extra amount of work. Actually, 1/3 of the candidates who were given the job offer came from universities that wouldn’t have been considered by conventional approach.

After the whole selecting process was over, we found out that more than 80% of the candidates hired had been recommended by our system at the initial stage of campus recruitment, proving the accuracy and predicting capability of our technology.