Why Genpact uses an AI chatbot to better understand its employees


Genpact cares about how its employees feel. So much so that it is linked to the compensation of senior executives.

Piyush Mehta, a thirty-year HR veteran, Genpact’s CHRO, leads the company’s HR practice using a suite of in-house technology tools that track and detect employee moods and feelings. The goal: to increase retention.

The professional services firm has a global workforce spanning India, US and UK and over 115,000 employees globally. With an average employee age of 28, Mehta is exploring how to better connect and understand such a large and dynamic organization. He spoke with Fortune on how Genpact uses an AI chatbot and organizational network analysis to better serve and engage employees.

This interview has been edited and condensed for clarity.

Fortune: Before I get into tech, tell me how you’ve seen your job in HR change over the past 10 years.

I’m admittedly cynical about a lot of the things said about what’s changed, having started in the profession 30 years ago. The reason people came to work even then, although it may not have been articulated as such, was the purpose of the company, the opportunities for growth and the quality of their work and their manager . This continues to be the most important reason people come to work today.

What has changed is the patience people have for these things. What has changed is the experience people want. What has changed is the technology that enables this. What has changed is data and analytics that have become an enabler to better meet the needs of our employees. But fundamentally the reasons why people come to work have not changed.

Can you tell us more about how Genpact leverages technology to assess employee engagement and connection?

Every business has a purpose and a culture to some extent. But these are just empty words if you don’t put them into action. Before the pandemic, the absorption of corporate culture happened by osmosis because people came to work, interacted with others and picked it up in an unintentional and unorchestrated way compared to what we do now. During the pandemic, there needed to be more intentionality in how we do things.

So we looked at three or four things. The first was, are people engaged? To measure engagement, we created an AI chatbot. To be fair, we did this before the pandemic, but we increased the upgrade to a different level during and after the pandemic, because it was extremely useful for us. His name is “Amber”.

How it works?

It is therefore an AI-driven chatbot that learns as it interacts with people. A regular survey didn’t work for us because you have to orchestrate it every time, and there’s a fixed interval by which it would act, which would then interact with the population and create results. These results would then be analyzed through data and analysis, provide insights, and then actions. But we needed to measure mood and sense of organization on an ongoing basis. Amber contacts employees in real time, at least four times during the year for the average employee and eight to ten times in the first six months for new hires.

Amber asks the employees questions. For example, he reaches out to me while I’m working on my laptop and says, “Hey Piyush, can I talk to you for a few minutes?” Can we discuss?’ And I chat. He asks, ‘How are you? What works for you and what doesn’t? At the end of the conversation, this generates a mood score for me based on learnings from previous discussions. The reason I say it’s AI driven is because Amber asks questions based on my previous answers, learning and getting smarter as she goes.

What are you looking for in the data and mood notes you receive?

Amber provides me with real-time intelligent analytics on the mood of the organization at all times and at the most micro level possible. I can view data by person and company-wide. It was huge during the pandemic. Response rates are around 65%. That’s phenomenal, considering he asks you questions four times a year. Someone who does is twice as likely to stay in the organization as someone who doesn’t. We like to measure the hell out of everything.

We also said during the pandemic that employee mood and sentiment would be extremely important, so we tied 10% of our CEO and Top 150 bonuses to Amber’s mood score. We changed our bonus pool matrix when the pandemic hit – in the middle of the year – and said it was so important that it would be the only non-financial measure of our bonus pool.

Do employees still have traditional check-in meetings with supervisors, or is this the new, more transparent way of doing things?

Basically, it’s always an employee’s most important point of contact, and Amber complements it. Think of our organization: 115,000 people and 10,000 supervisors. By definition, you’re going to have 5,000 supervisors above the median in their ability to manage their staff and 5,000 below. Amber allows them to better manage their staff.

The answers also go to the managers. How do you ensure employees respond honestly and don’t pretend to be happier than they really are?

There’s no pressure on people to respond, and about 35% of people don’t respond. If I’m not happy, I don’t answer. Because our positive mood score is above 85%, we see huge variation in how well we achieve this. Some are at 40%, and some are at 95%. As long as we continue to see variation, we know people are transparent.

We also have strict guidelines. You can’t victimize based on Amber’s comments because if you have a team of 10 people and one person feels victimized for their answers, they’re going to talk to the other people in their peer group and tell them to not respond, which will cause the response rate of this group. to fall. The manager is responsible if his response rates are lower than those of others.

What actions or policy changes has your team made in response to Amber’s findings?

We get a lot of career progression feedback from Amber. Employees felt that when placed in a position, they were required to stay there for 24 months before applying for other positions in the company. We didn’t think 24 months was long to be cast in a role, but a lot of people were leaving us. So, rather than losing these people to the outside world, we decided to give them the opportunity to grow internally in less than 24 months.

Genpact notably uses the analysis of organizational networks to assess people. Can you tell us more?

Organizational network analysis examines the metadata of networks within organizations. So it reads the subject – not the content – ​​of emails, examines the patterns, and derives analysis.

One application, which is profound because we are a high churn company, is the ability to predict churn before it happens based on specific metrics. This gives me the opportunity to step in and say that individual X is more likely to quit than individual Y, and I have to do the following things. There are data privacy implications, especially in markets like Europe. But we can do it in other areas because it’s within the legal limits.

What does your team do when Amber signals that someone is likely to leave?

We tell managers to integrate these employees into Genpact’s internal learning platform, so that they can grow. We also look at rewards and recognition data; when was the last time this person was rewarded, especially if they are reasonable? The third thing we look at is how long the person has held the position. If he’s been in the job for a few years and he’s bored doing the same thing, we move him to another job.

Amber Burton
[email protected]

Around the table

– The guy who inspired the phrase “quietly quit smoking” is back to working 50 hours a week. Business Intern

– Emotional intelligence was once considered unimportant to running a successful business, now leaders are obsessed with how to infuse it into their organizations. Listening time: 47 min. HBR

– The National Labor Review Board filed a lawsuit on Wednesday alleging that Amazon CEO Andy Jassy’s comments against unions may have violated labor laws. The Washington Post

– According to data from Glassdoor, employers are ghosting job applicants at a rate nearly double what they were before the pandemic. fast business

Water cooler

Everything you need to know about Fortune.

401(k) up. Employees may want to adjust their 401(k) contributions in 2023 since the limit will increase from $20,500 to $22,500 in 2023. —Alicia Adamczyk

Death of TikTok. A Philadelphia judge ruled TikTok was not responsible in the death of a 10-year-old girl who died while recreating a viral video that the platform had recommended to her. —Robert Burnson

Rejected, again. A railway union representing 6,000 workers rejected the last collective agreement, making it the second to do so since September. —josh funk


Comments are closed.