Over the past year, the Hewlett Packard Enterprise (HPE) human resources group has been using artificial intelligence (AI) and a chatbot to improve its talent acquisition and retention. And a company executive says the technology has turned things around dramatically.
Prior to rolling the AI-based chatbot technology, half of those who’d land on HPE’s career page looking for jobs would leave without ever applying. After the rollout in the first three months of 2023, HPE more than doubled the industry standard for visitors to its career site, reaching 950,000. The company is now able to convert 26% of “casual job seekers” into actual hires.
HPE is part of a wave of organizations taking advantage of AI tools for talent acquisition to supply HR shops with a full pipeline of candidates and to personalize the onboarding experience for new hires. Among the leading suppliers of SaaS-based intelligent talent aquisition platforms are Eightfold, Beamery, and Seekout — all of which perform skills inference and candidate-job matching. Other notable vendors include Clovers (with its recent acquisition of Talvista), HireVue, Pymetrics (recently acquired by Harver), iCIMS and Phenom, according to Forrester Research.
HPE chose to launch Phenom’s platform globally as its intelligent talent acquisition platform, but faced challenges ensuring algorithms meant to automate the hiring process were without baked-in biases and spoke to each potential hire in a personalized way. That process of customizing the hiring experience continues to occur — and continues to pay off.
Lavonne Monroe, who joined HPE’s human resources group at the beginning of the COVID-19 pandemic in 2020, is vice president of global talent acquisition and onboarding. Over the past year or so, her team has been leveraging AI and chatbots to create a customized career site that offers job prospects and current employees experiences tailored to their unique career paths.
As a pandemic hire, Monroe didn’t even enter an HPE office for the first year-and-a-half of her employment, so she understands and embraces the benefits of remote work, and the fact that it can be a talent acquisition manager’s dream.
Monroe is currently onboarding 67 new college graduates, 20 mid-level professionals and one new executive. Their views on hybrid and remote work were startingly different.
The following are excerpts from Computerworld’s interview with Monroe:
Can you tell me a bit about your role at HP, how that has changed with new technologies such as AI, and your approach to managing the workforce? “I manage all avenues of talent attraction, and that includes external and internal, and of course, onboarding to the talent. So, what’s their experience? I ‘own’ up to day five, but technically their first two weeks post-hire…, then we transition and hand them over to the hiring manager.
“From a technology perspective and how we’re managing our workforce, right now it’s all about retention and what tools are we utilizing for our team members navigate their careers. I look at it from a scope of lead generation — where you don’t even know if you want to work for HP because you haven’t thought about us — all the way to retirement; how do we propose that path for you in a seamless way, delivering a customized and relevant experience at the right times. We all know we have different life transitions, where we’re open to having more responsibilities at work or not…depending on our life transitions. How do we as an organization support all of that using our tools?
“That’s really what our talent management team at HP has been focused a lot on in the past year.”
What kind of change have you navigated over the past three years? “Unless all the companies come together and say the same thing, there’s no way they’re going to get everybody back in. They’re trying to find a happy medium. Even here we’re doing a hybrid model; if you’re within a certain radius of the office, you know, come in two or three days a week. Just pop in and get to know your team members — their face and names. That’s the piece of the remote aspect that we lose – the people connection.
“Our journey has gone from figuring out a 100% virtual world — not because we wanted to but because we had to — and now we’re going back to exercising in-person muscles that we haven’t exercised in a while. We just did a couple of in-person hiring conferences and they were a bit clunky for us. We haven’t done this in a while and we forgot how to recruit in person. It was really pretty funny; we didn’t bring extra shoes; we didn’t have a QR code. Right when we got good at the virtual piece, now we’re popping back and going to in-person. It’s really just ebbing and flowing with the market and knowing that whatever I put in place today, possibly won’t be something we utilize two years from now.”
You said via email to me that HPE’s career site offers candidates an “Amazon-like experience.” What does that mean? “Whenever you go to Amazon…, based on your history, there are recommendations that come to you — consistently now, no matter what browser you use. What I partnered with Phenom on is when a candidate comes to our career site, I want them to have an experience that is of them. What I mean is when we rolled out Phenom, we established landing pages by [demographic] region, and we intentionally established landing pages by job family and discipline. If I’m a software engineer in India, when I hit the landing page, I get an experience that I can picture myself in.
“Then the chatbot secondarily has assisted us in being able to communicate instantly with applicants versus the old-school recruiter phone call and … chasing them and having phone tag over several weeks, while not even knowing if [they] want to talk to HPE, and then never coming back to our career site. The implementation of Phenon has really helped us in garnering that kind of candidate experience, that personized, regional touch. Then, the chatbot is the instant gratification. I have a chance at communicating with somebody and get answers to just the basic questions I have without having to chase down a recruiter.”
How have you deployed Phenom’s AI-based HR tool? “We rolled out Phenom globally for all of HPE. It manages our career site. Obivously, [we use] the chatbot. We rolled out their apply process, their candidate management tool – so the CRM, as well. We’re in the process of turning on the event’s scheduling aspect of it for university utilization. We’re also looking a bit more at some other AI, but we’re also going through some legal hoops that take more vetting for us. We’re a bit more passive when it comes to the new tech and all the laws coming out now. We’re just a little more hesitant jumping out there any being first with the AI utilization, specifically on candidate scoring. So we did not turn that on. We did turn on pretty much everything else.”
New York’s Local Law 144 and the EU AI Law could become an international default standard like GDPR. What steps are you taking to preemtively address the issues of bias in AI tools? “We actually put together a small tiger team and we’re actively talking about this now, because what we’re finding is ‘yes, they are [biased]’ and we’re understanding this is a technology that has to be trained and fed and utilized in order to get good. There’s still a person who’s feeding that, so that could present bias.
“One thing I’m realizing is we have to have a way of auditing our algorithms, and that is not a standard job function within a lot of our organizations. Right now, what we’re talking about is whether we need to bring this kind of knowledge in-house so that we can turn on AI in a lot of different job families and functions, not just in HR. Or do we want to partner and find an outside legal counsel or organization. So, it’s not a matter of not doing it, it’s a matter of how we’re going to go about doing it, and then how we’re going to go about auditing the algorigms and ensuring we are fixing any findings and we’re able to catch on fast so that we don’t have bias within our hiring process or internal mobility process. We want to be able to use it for recommending internal jobs to our team members.
“How can we legally do this and then what resources do we not necessarily have that we may need to go ahead and hire for? So, it’s a different skill set we’ll have to go out and look for potentially, if we want to bring it in house. We’re right at the beginning of that.”
HPE was able to convert 26% of casual job seekers into actual hires. How did you do that? “Before Phenom, our career site was managed by our RPO (recruitment process outsourcing) vendor, and we did not have the foot traffic to our first site. We definitely didn’t have a chatbot experience. As soon as we turned that on, our career site traffic literally doubled industry standards. So we had like 960,000 site hits within a three month span, and that’s largely due to those [customized] landing pages we created and how we were driving candidates to our career site. The chatbot experience gave us the opportunity to speak to people that were dropping off the site.
“Before, when they came to our career site, they were staying for maybe 30 seconds to a minute max, and then they would drop and never apply. The chatbot was helping us with keeping them and eliminating the need to go to multiple pages on the career site to find information.
“And, then the chatbot is also delivering them the specific job to apply for. So, it’s giving them the lead directly. So, they’re clicking, they’re applying.”
What were the results before your career site upgrade? “We didn’t have the chatbot before, so we were literally losing nearly 50% of our career site traffic; we weren’t converting them to apply to applications.”
That’s a huge turnaround, if you were losing 50% before. “Yes, and they’re staying longer, so they get to experience and learn more about us. But it is mainly due to our chatbot versus clicking on multiple pages on our career site. That’s the other key thing we’ve noticed. Before, for example, we were just delivering them a page for benefits, a page for who HPE is, a page [for this or that]. You had to click the page 10 times. Now you just have a conversation with a chatbot.”
Was the chatbot organically good at bringing candidates the right responses and leading them to the right jobs? “That’s the other key thing about this that’s similar to AI. We have to feed. We have to feed the Q&A; we have to train the chatbot consistently. That was another part of the implementation. I would say the key to success in our implementation is we got business leaders involved, HR involved, recruiters involved, we even pulled in recent new hires that came to HP without the chatbot experience and we had them help us with the Q&A to establish the chatbot’s ability to answer questions. That’s always going to be a work in progress. We’re always loading information and getting better.
“We even asked them, ‘What made you stay on the site?’ Because before Phenom, our average apply time to get through a job application was 15 minutes. Once we turned on Phenom with their instant apply, it went from 15 to five minutes. We’re in the era of instant gratification, so time is golden. That was a great win for us.”
How have you created a recruitment pipeline that’s both culturally sensitive and regionally specific to candidates? “Phenom’s tool has allowed us to create multiple landing pages so that we can get that experience. So, I see Phenom as the tool helping us do that versus before when we had a very static career site.
“But from the aspect of a candidate’s experience, it was clearly understanding that engineers want to be talked to like engineers [and] knowing that an HR [applicant] doesn’t want to hear a lot about engineering. It’s understanding the behaviors and creating career profiles aligned to our key critical job families we want to recruit for.
“When I came in, I looked at all the historical recruiting numbers and said, ‘OK, where are our critical jobs, what’s driving revenue for HP, what’s keeping my CEO up at night if I have open acquisitions?’ And, those were engineers, product managers, sales, and then some functions. So, what are my candidate profiles aligned to an engineer, what are the behaviors of an engineer, what does an engineer want to see? It’s really going into the marketing aspect of recruiting and then delivering that experience that is relevant to that candidate, and not forcing an engineer to go through an HR-friendly career page that doesn’t even talk about a technical career path and how they can be a technologist or an engineering manager.”