Ceridian made headlines in 2012 when it acquired Dayforce. The acquisition eventually extended Ceridian’s offering to include a work hub, automated tasks, and even a wallet that enables flexible payments from employers.
The impact of these functions on the world of work was illustrated earlier in December at Ceridian’s London stage of its World Tour as Costa coffee revealed they had managed to retain all their staff and enable flexible working during the COVID-19 pandemic.
However, I was keen to hear more about the data and Ceridian’s technology itself. Fortunately, David Lloyd, chief data officer (CDO) at Ceridian, was happy to sit down with UNLEASH (in real life for a change) and discuss Ceridian’s offering, all things data, and how artificial intelligence (AI) can impact every part of the employee lifecycle.
Capturing data and recruitment
Lloyd explains where the data that Ceridian uses is kept: “A lot of the employee data that we capture within Dayforce is centralized. So we have customers in the Azure site with Microsoft’s cloud offering there, and we also have a group of clients that eventually will be there, who are [still] in they’re our own private hosting environment.”
Nonetheless, the company intends to move most of its clients, and their subsequent data, to the Azure cloud in the near future.
Ceridian uses AI across multiple products in its human capital management (HCM) suite that enables the company to leverage the data it has in the cloud. To get a better picture of how this AI and data is used in conjunction, we should begin at the starting point for employees: the hiring process.
Lloyd notes: “Basically, it starts with the premise that the algorithm starts to learn about the best match between the candidate’s profile, and then specific job attributes, and then as a recruiter or manager corrects that, which they can do in the platform, it learns from it.
“We’re very cognizant of something called trustworthy AI.”
The reason Ceridian is conscious of trustworthy AI is: “In the UK [and the] EU, the trustworthiness of artificial intelligence is a big thing.
“So we have to make sure not only that the intelligence that we’re using is explainable and that people understand how it works, but more importantly, that we’re doing things the right way.
In terms of the hiring process, AI can strip out features like “name, address, ethnicity, sex, and age” as well as elements like which school someone went to from a CV.
Additionally, the AI often disregards “data in your resume that’s older than five or six years old, because really [when] you’re talking about the skills economy and your most current skills are the ones that matter most.”
From this point, Ceridian is “able to let the system look at a grade collaboratively with the recruiter and the manager, and then they can tweak it accordingly”.
Lloyd adds: “A lot of that early filtering, and the recommendation part, is done without introducing human bias into the process.”
Bias is an ongoing issue within many organizations, as minorities leave jobs at an alarming rate in the ‘Great Resignation‘.
Data in onboarding and beyond
Evidently, the issues of bias extend beyond the hiring process, and Lloyd went on to tell me how data can be used to improve the experiences of employees throughout their journey.
On the topic of engaging employees, Lloyd comments that the best thing to do is “create what I call low employee effort.”
Lloyd unpacks this term by explaining that it requires making it simple for retail workers to swap shifts or learning management systems giving employees the answers they need quickly and effectively.
Equating a good recommended answer process to Netflix‘s approach to recommending what film you should watch next, Lloyd notes that suggestions come from data collected from you and your colleagues and can then be quickly available as well as accurate.
Like anyone who has also run into a useless PC diagnostic tool or Googled symptoms to discover they are facing immediate death, I was keen to hear how employee experience suggestions and answers could adapt.
The answer is that Ceridian’s HCM system can leverage data from when an answer is marked as unhelpful or a suggestion isn’t used. This feedback loop is essential to the learning of the tool.
Lloyd adds: “That’s one example. If you’re doing it in other areas, such as benefits where we use intelligence to recommend [offerings] because you and your partner have two children, or you have a certain income level, we know what you’re making, we can better look at what we would recommend for the best benefits fit for you, based across a number of different benefits.
“But if you don’t like that, you may say I don’t want that benefit, I want this one, and the system can learn from that.
Lloyd continues: “A lot of it [the AI solution] is based on an interaction with the employee who starts to train them.”
“The other thing that’s really key, is that you also have to have a really good set of data that you’re training against, and an exceptionally good set of data to test against, and they are not the same.
“So using that and taking that time, it means that you should be testing the systems very regularly, to make sure that bias isn’t creeping into the model.
“The only way to do that is to be diligent about a regular testing regime and carry out consistently across the model to make sure it’s doing what you expect.”
Of course, testing the system requires ownership, but the benefits are clear as Lloyd talks about the experience employees have and want.
“You can look at HR problems differently. I think a lot of how we need to think about all of this is consumer experiences, HCM systems were not built as if they were consumer experiences.
“They don’t work the way Netflix does, or the way Snapchat does, or WeChat, or the way that you may shop at Amazon or look for your favorite videos on YouTube.
“But arguably, every one of us as employees, absolutely expects the consumer experience [at work] now because that’s what we have in our pockets all day long. “
Although there is a clear employee experience benefit of Ceridian’s approach, it will not be enough to retain every single employee.
Leaving a company
Lloyd discusses how data can be used to better inform companies about why people are quitting, a particularly important topic at the moment amid the ongoing competitive war for talent in the US and across Europe.
“When you see people leave, you’ll be able to go back and say, ‘How many times did they speak to their manager in the last three months? What was their last set of compensation changes over the last three years? How were they recognized for their work? What was their performance? What was their career path and talked about career paths?.”
He adds: “All that data can now be linked to [whether] there’s a rationale for that person leaving”, such as “I got a better opportunity for more money”. “And now that you have the rationale, you can then look back at the data that helps us not necessarily support that decision. But understand how that decision, we ended up coming to that.”
Conversely, Lloyd stresses that data can be used to find out who top performers are, even if that is not evident in traditional ways of monitoring employees. This can help organizations retain staff who may be tempted to leave because they feel undervalued.
Evidently, data can play a huge role in the employee lifecycle, and Lloyd made it clear that this kind of investment with the right ownership can help to retain and attract the best talent.
On the back of this interview, the final question then rests with the reader, will you invest in data?