The biggest enterprise AI story of 2023 is not generative AI. As I documented at CCE 2013, it's all about data (CCE 2023 - the great AI debate shifts from content creation to...data governance?)
The caliber of your data is the deciding factor on your analytics, your "classic" AI, your generative AI, and whatever shiny new toy comes next.
No surprise, then, that a standout use case of 2023 was how Accenture's skills ontology changed its approach to talent (Workday AI/ML Innovation Summit - how Accenture shifted to a skills-based organization with Skills Cloud). The one missing piece from my time at Workday Rising US? Building on that story. But last week I had the chance to do just that, via a virtual chat with Unum Insurance on its Workday Skills Cloud pursuits.
When you talk about cloud financials and cloud HR at scale, Unum fits that bill. With over 10,300 employees, Tennessee-based Unum insures 39 million people worldwide. Unum also provides well-being and leave management services. Prior to moving to Workday Enterprise Management Cloud, Unum was running 39 different financial and HR products (Unum has now narrowed retired 20 of those disparate apps, aided by their Workday Financials go live in August 2017, with Workday HCM already in place). The Unum use case on Workday.com documents notable results, including "$200K net annual cost savings, and $1.7 million annual head-count cost savings."
Why Skills Cloud? "We're always trying to look ahead a little bit"
But how does Skills Cloud fit into the picture? As Lisa O'Brien, Director, Global HRIS at Unum told me, Unum considers Workday its go-to-platform:
Unum is a full-platform Workday customer. We're really looking at all of the different modules within Workday, and trying to really build on our investment there. In my role, I'm looking at the HR side of the house. So we're using core HCM - Recruiting, Learning. We're doing all of the reporting capabilities, we're also using things like Prism and Extend. So we're always trying to look ahead a little bit - like what else is coming down the pipeline that may benefit us, and may benefit some of our initiatives.
Skills Cloud fits that bill:
When we first heard about Skills Cloud, there was a lot of interest. This was about three years ago, but even then, there were rumblings of 'This is the hot thing,' right? This is where every organization wants to be - the skills-based workforce. Using skills is the currency to connect employees to opportunities. So we aspirationally wanted to start moving in that direction.
But skills ontologies are not necessarily easy to build or manage. Prior to the AI surge, abandoned skills projects were common. With 10,000+ employees, Unum's Skills Cloud had to scale. Could Workday make this a viable option for Unum? O'Brien:
From that lens, having Workday Skills Cloud we knew would be beneficial, because we wanted to tap into that big repository of the skills inventory, and have Workday do some of the heavy lifting for us in terms of maintaining it.
To pull this off, employee buy-in was essential - not to mention an intuitive self-service interface. Assistance with data cleansing was also needed:
Skills was a bit of a wild landscape before we turned Skills Cloud on, because we had a lot of employees entering all different types of things. But with moving to Skills Cloud, you get the benefit of Workday doing some of that maintenance for you, and looking at duplicates or synonyms. And you're tapping into a much more robust inventory than we would ever be able to maintain otherwise, so there's a benefit there. [Author's note: Unum went live on Skills Cloud in February 2019].
Why does a modern skills ontology matter?
In the past, the downside of building a skills ontology kept many such projects at bay. But this is clearly changing: skills projects are getting easier, and the payoffs to a dynamic, well-maintained skills ontology are eye-opening. O'Brien:
We also know that the Workday Skills Cloud was the entry point into some other features we thought would benefit our workforce, such as Career Hub and Talent Marketplace. We went live with Career Hub about two years ago. Skills Cloud is the underlying infrastructure beneath Career Hub, and then also with machine learning.
There are a lot of really beneficial features within the Career Hub that benefit employees and their development. So skills, machine learning - it all kind of comes together to give something back to employees - and lets them take charge of their careers.
Let's talk about that. What benefits are employees seeing from this effort?
Employees feel strongly that they need to have access to career development resources. They want to feel like they have a career path that helps drive positive employee engagement... If you take away the Skills Cloud and Career Hub, then what do employees have? They have an open job listing. But how does an employee know if they're a good fit for a role?
What's great about taking Skills Cloud and the Career Hub back into the equation is: with that machine learning, Workday is making recommendations of, 'Hey, here's a job that you might specifically be a good fit for, based on your skill set.' Or: 'Here's a skill that you said you were interested in. So here's a learning that you could take, that would help develop that skill.' Or: 'Here's a potential mentor that you could match with, who has said that they're an expert in that specific skill.' So all of that framework is available.
O'Brien's team still has internal adoption goals - but getting the system up and running was not a barrier:
Today, we're working on adoption and taking it to the next level. With very little configuration needed as a customer, you can get all of those benefits that machine learning is doing. But we can also help to make the machine learning better, if you will, by continuing to turn on more features.
Right now, Workday's Machine Learning will make recommendations for different learning courses, based on descriptions of the courses, titles of the courses. But we don't tag all of the specific skills to our courses yet; that's on our roadmap to do. If we tag a specific skill to a course, Workday's machine learning will get even better. It's very good now, even without the tagging, but it will get better. And we'll be able to put forward certain learnings that we want to prioritize, based on that tagging getting accounted for in the Machine Learning.
Impact on mentoring - "intelligent" matching can change careers
Of all those capabilities, mentoring really jumped out. If Workday can help a customer like Unum make more pro-active mentor matches, it's hard to overstate the impact. I think back to the mentors I was lucky enough to connect with at the right time. How hard would it be to find such mentors today, in our Zoom-and-run workforce? I asked O'Brien: any mentoring successes so far?
Unum puts a big emphasis on mentorships. There's certain areas of the business where mentors are basically required, as you work up through the ranks. There's a lot of structure in those business areas to try to match people with potential mentors. But there's certainly a big part of the population where there's there's a less formal structure around mentors. So how do you give those people the same opportunity to find people and connect with them, and benefit from that mentor relationship?
So the Career Hub and Skills Cloud definitely allows us to tap into that. Employees can not only enter what skills they have; they can say what skills they're interested in. And then on the flip side, potential mentors - just with some of the mentoring capabilities within Workday - can say, 'Here are specific skills that I would be comfortable mentoring on.' So it definitely helps to match up those opportunities.
The wrap - skills are a discipline, but the payoff is real
If you want to shift to a learning/talent-centric organization, a modern skills ontology is non-negotiable. That doesn't mean these implementations are easy, or that every employee will immediately begin using the system.
But Unum's story shows that these projects are now within the scope of what's possible. Workday launched Skills Cloud in 2018. Now, out of 4,000+ Workday customers, 1,500 are on Skills Cloud - with five billion skills in use. For Unum, the 85% adoption of Workday performance tools is a promising sign for adoption in skills as well.
Yes, it takes time to prove out the ROI on some of these talent-oriented metrics (Workday makes the case in Does Your Approach to Skills Management Strengthen Upskilling and Employee Retention?).
Precision matters: AI, in all its probabilistic flavors, is not going to be near 100% accuracy. Therefore, personalized training and skills recommendations made by such a system are not always going to resonate. But I don't believe that should be a deterrent.
As I said to O'Brien, even if the "AI" sends you training or mentor recommendations that aren't spot on, as long as they are in the ballpark, those kinds of recommendations can provoke a vital conversation. This might lead an employee to ask about a different mentor, to update an out-of-date skill, or to add a new desired skill to the Skills Cloud. O'Brien says that's exactly what's happening at Unum.
Just speaking from experience as a hiring manager in the organization, anytime I post an open job, I get so much interest from internal employees, but they want to know: 'How do I become a good candidate for your opening?'
That's where Career Hub comes in:
I keep saying Career Hub, but you can't have Career Hub and Talent Marketplace turned on without the Skills Cloud. The Skills Cloud is foundational.
O'Brien is also signed up as a mentor within Workday; she regularly hears form employees looking for guidance. That is fuel to take the promotion of these new talent systems to another level internally.
I am also watching HCM vendors to see if they can further automate skills ontologies, to make them even easier to maintain - because for AI, the currency of that data is everything. Could Workday, for example, further automate the "smart tagging" of skills as employees onboard, and move into new roles?
This isn't as easy at it sounds. Many companies - Unum included - have custom skills that are not simple to add and track. O'Brien says she is having these kinds of discussions with other Workday Skills Cloud customers, as part of the Workday-led Skills Cloud advisory council. That kind of frank customer/vendor dialogue is at the heart of the results we all want to see.
The Skills Cloud data is pulling Unum leadership into the mix. They are asking about talent in new ways:
We're getting a lot of positive feedback around being able to identify skills and skills gaps differently. There's still more we have to do here; we don't have it all figured out. But there is this universal understanding, I think, that this is where we need to go as an organization to continue to be successful, and to stay competitive.
Definitely a story worth tracking.
End note: diginomica colleague Phil Waineright is back from Workday Rising EMEA and is posting notable updates and analysis, including Workday Rising EMEA - 'We're the team that shows up.' Workday's push for regulation to close the AI trust gap.