This growth (which has doubled in the last three years) has been made in the face of turbulent economic times, with global events impacting hiring trends and activity in dramatic ways. Throughout this, the business strategy has centered on developing sophisticated new technological solutions built on artificial intelligence (AI) and machine learning (ML).
As I learned from Raj Mukherjee, Executive Vice President and General Manager for Employer at Indeed, when he joined me for a podcast discussion, the focus of this has been on understanding and catering to the individual seeker and employee requirements to create more personalized and successful connections.
So, let's take a look at how Indeed has been using AI and ML, as well as dive into some of Raj's ideas about how it will go on to transform his industry (as well as many others):
The Career Companion
Career Companion is a phrase that Mukerjee brings up during our conversation, and it serves as a perfect description of Indeed’s AI ambitions. Broadly speaking, the business objectives fit into over-arching categories. These include aligning the service more closely with user preferences or improving the speed of its processes.
Mukherjee tells me, "Our founding principles were around machine learning.
“We use standard machine learning techniques, we also use deep learning …and now we’re leveraging … generative AI or large language networks.”
The career companion concept stems from developing the understanding that an end-to-end employment platform isn’t only going to be useful to candidates one time when they're looking for a first job. Today, they might be looking for a job, and tomorrow it will be a promotion. After that, they might start to get involved with hiring staff themselves.
Understanding preferences is key to this – Indeed collects more than 140 million data points every day, which is calls "qualifications". These are used to develop an in-depth knowledge of applicants and how they match up to roles advertised on the platform.
Describing how AI technology has enabled new possibilities for using this to drive customer experience, Mukherjee says, "We might have a resume that I've shared with you, and that resume has lots of data about me, my past experiences, where I worked, what type of skills I had in those.
“We are going to understand that. Using parsing technology, but also then starting to use machine learning at a very deep level to extract the right skills.”
Speed is also of the essence. Because of this, metrics around how quickly it can fill the vacancies are among the most keenly monitored.
Mukherjee tells me, "We know employers want to hire fast … if you use our paid product, it leads to 19% faster hiring.”
This is down to the fact that roles are actively matched to applicants via AI, who are then invited to apply as soon as the vacancies are posted on the site.
Behind the scenes, however, a lot of complicated matchmaking is going on. Indeed’s algorithms unearth insights based on both “stated” preferences and unstated preferences. Stated preferences are provided by the user – such as an indication that they are interested in vacancies in London or Paris.
An unstated preference, on the other hand, is one that is determined by analyzing the data. For example, if they are consistently searching for jobs around London, it might determine that it’s where they want to work.
Combining these stated and unstated preferences has been found to efficiently speed up the critical time-to-fill metrics.
"We can start to create a picture in the machine, and then that will lead to matching … to better opportunities”, says Mukherjee.
“The same thing happens on the employer side because we have billions of job descriptions that we’ve extracted over the years.”
The Personalized Agent
With its strident investment in building AI into the core of its service from day one, Indeed has its sites set on providing job seekers with their own fully automated "personalized agent."
This will help to improve one metric that Mukerjee particularly hates. This is because when applying for jobs, three out of four hopefuls simply never hear anything back.
He tells me, "Even a rejection, while not great, is better than not hearing back.”
One simple change emerged as the clearest route to reducing the number of times that job seekers would experience that frustrating outcome: Reducing the number of jobs they have to apply for before being hired!
“So how do we get to a world where job seekers don't have to apply to so many jobs? They can come in and have … I call it a ‘personalized agent’.
“So … you share 140 million preferences and qualifications with a machine, and that machine understands you and returns back jobs that you have close to a hundred percent chance of hearing back … As an industry, we are failing every single individual who applies for a job and never hears back."
The Future of AI in Hiring and Recruitment
As Mukherjee clearly has deeper insight into the changing world of work and the shape of the future job market, it is sobering when he says, “many jobs will disappear.”
"It's a fact," he tells me, "It has happened before; every technology revolution has led to loss of jobs.”
"But new jobs get created, and there will be many new jobs created as part of this AI revolution.”
Those caught in the middle -neither made redundant nor finding themselves filling entirely novel roles such as prompt engineer or AI ethicist – may find themselves in a brave new world.
In this world, AI will take care of matching, determining connections, and automating the mundane elements of the hiring process.
We, on the other hand, will be active in the “human” aspect of the HR role – using skills such as our empathy, communications abilities, and capacity for creative thinking to create outcomes that advance the best interests of everyone in the business.
“I do believe … that what will end up happening is all of us will enjoy our jobs more,” says Mukherjee.
“We will do less of the stuff we don’t want to do, and we will do more of the stuff that we want to do. That’s the optimistic or realistic side of me speaking, and I do want to make that happen.”
You can see my interview with Raj Mukherjee here, where we take a deeper look into more of the transformative implications of AI meeting HR.