Google Job Search API

Is it a game changer? What does it mean for Indeed or Glassdoor?

Ivan Davchev
3 min readMar 11, 2017

I just came back from the Google Cloud Next 2017 conference. There Google presented the Cloud Jobs API and their roadmap for upcoming features.

A couple of years ago this work was still in stealth mode at Google, with an opaque code name, much like the Invasion of Normandy :)While heading job search at Glassdoor at the time, I met with some of the Google team behind this project. It’s great to see it taking form and becoming available outside Google.

What Is Google Jobs API?

It is a turn-key solution to searching your jobs. Like most cloud-based search solutions, you need to feed it documents (jobs) which it indexes and makes searchable.

The awesome sauce is in query processing. Google has analyzed millions of jobs and applied state-of-the-art machine learning techniques to recognize and extract skills, job titles, and the relations between them.

It knows, for example, that “SDE II” means “Senior Software Engineer” and can also be “Java Ninja” at certain companies. Google is historically excellent at this — and I worked on such query processing features at Google years ago, so my biased opinion comes from first-hand experience. There are good reasons to believe it is at least as good at query processing as the leading competitors in job search — especially since I happen to have worked on their search too :)

Neural nets and deep learning. Soon, singularity ;)

Another great feature is smart location search. It is aware that simple radius distance is not good enough, especially in places like New York or San Francisco where bridges and tunnels make commuting complicated. Google demoed commute-time search (e.g. “I don’t want to commute more than 30 minutes”), which is apparently coming soon. For now, location is used to inform ranking the job listings.

What’s Missing?

It does not yet use behavioral signals (clicks or applications to jobs). So it is a very fancy text search engine tuned for jobs content. In a future iteration they say they will incorporate behavioral signals.

It is not an aggregator and does not offer “all jobs in the world” out of the box. So you can’t build an Indeed competitor in an afternoon with this API. It only searches what you provide it, which is why it makes sense for job boards and applicant tracking systems (ATS) to use it — jobs are posted on and managed by those systems.

No sponsored or promoted jobs yet. Something along these lines was on the roadmap. Another reason why you can’t build an Indeed with this. All jobs are treated the same: either all paid listings (job board use case) or unpaid organic (employer job site or ATS use case).

Pricing: unknown. The API is still closed alpha.

What’s Next?

Here is a roadmap screenshot. The resume-to-jobs recommendations is interesting. The candidate resume becomes the search query. Others have tried to do this with limited success, including my former teams at job search companies. My hope is that Google can crack this difficult problem and provide a great experience.

Is It a Game Changer?

Not yet.

It is transformational for an employer’s career site. Johnson & Johnson is experimenting with this API (pre-production) and they spoke about the big improvements in relevance they saw on their career site.

It is not going to change much for job aggregators like Indeed or Glassdoor. In particular, Indeed does have smart location, related job titles, behavior-based ranking, and behavior-based recommendations. In addition, small businesses which do not have applicant tracking systems can directly post their jobs on Indeed for free, and these jobs are unlikely to show up on Google Jobs API soon.

When Google adds behavioral ranking, recommendations, and promoted jobs, it will enable interesting use cases which can lead to bigger changes in the industry. I am looking forward to seeing how this API evolves!

--

--

Ivan Davchev

Engineering leader, search expert, data scientist, cryptographer, and (bio) hacker. Leadership roles at Amazon, Google, Microsoft, and startups.