A big first step to re-inventing job search

Sjoerd Gehring

Recruiting at Apple

Published Mar 19, 2017

+ Follow

Last week during Google Cloud Next ‘17, Google profiled its new Jobs API. Johnson & Johnson is proud to be one of the first companies to bring a new way of job search to our candidates in close partnership with JIBE and Google Cloud.

Machine learning-powered job search

Everyone is talking about artificial intelligence and machine learning these days but how can these powerful technologies help improve the candidate experience? The reason we love Google’s new Jobs API so much is that it uses machine learning to power our career site with a much smarter job search. The Jobs API uses its understanding of language to return highly relevant results to complex queries, even when the exact keyword isn’t part of the job title. This way, Google Cloud machine learning helps job seekers not only save time by finding the most relevant jobs, but also discover opportunities that otherwise wouldn’t have surfaced.

In the example below, a job seeker might be discouraged with the top search results on the left (with our previous keyword search) returns job titles that have no mention of Oncology even when the search query was “oncological research”. The results on the right side however, return many more relevant results. The Jobs API machine learning model is structured to identify the relationship between many items of career relevance – skills, education, companies, seniority and location. This results in surfacing more relevant jobs for the job seeker.

At J&J, we’re on a journey to re-imagine the way we identify and attract talent. We’re proud to work with Google Cloud and JIBE on this initiative. I’d love to get your thoughts on this new enhancement to our candidate experience and I hope you take a look on our career site and share your feedback.

Lisa Schwartz

Kudos on continuous process improvement. And, here is another area of opportunity. When a candidate applies for an open role at J&J, they receive a survey one month later even if they never spoke to anyone at the company. The survey asks , "How likely are you to recommend Johnson & Johnson as a prospective employer to family and friends?" How could a candidate answer that question any better after hitting the submit button? If J&J is interested in measuring how they are "perceived" by candidates, then that question should be asked at the time of application. If J&J is looking to improve their application process, then they need to ask relevant questions. The next 2 questions should be automatically captured via J&J's application process. "Did the recruiter contact you and provide a reason why you were not a good fit for the position you applied to?" And, "Did the recruiter invite you to join Johnson & Johnson's talent community?" Any CRM app can capture this data once a recruiter make a phone call or sends an email. J&J asks Question 1 again with a slight twist, "How favorably do you view Johnson & Johnson?. I assume if someone is applying for a job where they're going to spend about 2000 hours a year, I'd hope they view that place as favorable. And, "To what extent has the recruitment process influenced the above score?". Again, this is an application process, not a recruitment process so not sure hitting submit can influence someone one way or another. And, the final question, "What can Johnson & Johnson do to improve the recruitment process?". The answer? Ask meaningful questions of those taking the time to fill out a survey. Maybe you would like to know the top 5 reasons people apply to the company? Is it the dental plan, work-life balance, diversity, or the number of women in leadership and BOD positions? Monitor and measure as much of your process in stealth mode; the best processes are those that are invisible. And, think about potential bias in data gathering . That is, J&J is asking job seekers a broad question about how favorably they view the company? Even though the survey is anonymous, folks may be inclined to provide positive responses with the hope there is a future job opportunity. Ask more specific or targeted questions.

LinkedIn User

Very exciting how technology con simultaneously help job seekers and employers to find the best fit! --It would also be nice to be able to use more filters- for example I don't see on your career site where you can filter years of experience. This, at times, can also be discouraging when the first 5-10 results have requirements far out of your reach.

Tracey Parsons

Great work! I love that whenever you publish it is about making the candidate journey more enjoyable and friction-less. Kudos to you and your team!

Loïc Fourot

As employer, also take a look at Riminder.net. An AI solution to evaluate resumes quickly with exceptionally accurate Deep Learning.

Nancylee (Nancy) Matles- Certified Executive Coaching, Digital Marketing

Jay, Thanks for posting. Today AI is not able to replace a real live Coach who can evaluate options, identify "The Right Fit" and help candidates negotiate offers to land new positions. Nancylee Matles

See more comments

Experience vs Education. What counts for more when hiring?

Nov 20, 2017

Here’s a sobering stat: employee disengagement is at nearly 70%...

Nov 10, 2017

Failure isn't always a bad thing...

Nov 8, 2017