Main content

HR in an RPA vendor – how bots are transforming the function

Brian Sommer Profile picture for user brianssommer January 5, 2021
A conversation with Nancy Hauge, CHRO of Robotic Process Automation (RPA) vendor Automation Anywhere.


Many CHRO interviews are pretty pedestrian. Most involve an executive telling me about their new HRMS solution and the implementation effort they incurred. Some particularly painful ones occur when an HR executive wants to prattle on about some minor item (eg: “Brian, did you hear, we changed the color of our logo? It’s really driving up our engagement scores!”). You can’t imagine what a relief it is to speak with a great, business-oriented HR officer who’s really doing something value-added, innovative and newsworthy.

I recently had the opportunity to speak with Nancy Hauge, CHRO of Robotic Process Automation (RPA) vendor Automation Anywhere. One of the headlines from this interview was that it highlighted a number of interesting ways HR and the enterprise can use bots. Lots of bots.

How many people work here?

This problem still frustrates executive teams everywhere. Why? Whatever headcount numbers that HR can produce, the odds are good that it won’t agree with Finance’s numbers (from Payroll or Budgeting) or Operation’s headcounts. The HR system only contains active employees but may not be able to calculate a full-time-equivalent value when part-time employees enter the equation. One client of mine has a number of ‘alumni’ (ie: former employees) that they re-activate for short-term needs (eg: they can reactivate a nurse in a given location if the regular one is on maternity leave). Are these people employees, part-time employees, contractors or some other status? How do you account for them? Ditto for interns, some contractors, people on sabbatical, people on extended leaves of absences and other worker types.

It’s fascinating to see how employers conduct one of these employee census exercises. The usual method involves a lot of emails, faxes and spreadsheets. Some work may be required to normalize and cleanse the data as some employees may work at two different locations and get double-counted (this is incredibly common in the restaurant & convenience store world). The quality of the results is often compromised as front-line managers may not count employees consistently or correctly.

A bot was a better way to go for Automation Anywhere as:

  • The bot could integrate to the various HR systems
  • The bot could consistently apply the categorization and counting rules across the firm
  • The automated bot could read data from spreadsheets, payroll reports and HRMS systems and would automatically map the correct data every time the census was needed
  • The bot could do the job in a few minutes while the manual process took days or weeks to complete
  • People came to believe the results from the bot while the manual process still had naysayers

Who’s eligible for benefits?

For many employers in the US, employees who work 30-hours or more a week are eligible for benefits (eg: health insurance). For many employers, they want scheduling tools to cap work hours to just 29 per week (a bot could help with this). But, whether the firm can control this or not, there still is the business need of determining who is eligible for benefits. And, this determine usually must occur after each payroll run.

The eligibility issue is complicated by a part-time employee having coverage through their spouse’s employer. So, for HR, the issue of eligibility is two-prong:

  • Has the employee been working enough hours to qualify for benefits?
  • Will the employee want to sign-up for benefits if they are deemed eligible to do so?

For Automation Anywhere, a bot was able to determine eligibility in just 40 seconds while the previous process took around 9 hours.

Bots as part of HR implementation

Another interesting use case for bots in HR involved their use in the implementation of the HRMS itself. Automation Anywhere used bots to move data from their old systems into their new Workday HR applications. That wasn’t insignificant as they’re converting several local solutions to a global cloud product. Bots were grabbing data out of several, different systems and formatting it into specific file formats for the Workday application.

But, what was even more interesting was that the bots were used to convert unstructured data, too.  So, items like employee photos, proof of identity documentation, tax exemption requests, court orders (eg: for garnishments), etc. were all captured and uploaded.

New takes on the 9-Box – HR for Individuals

In HR land, there is a 3x3 matrix (aka the 9-Box) that helps HR/Management see how well people perform (on one dimension) versus what they are paid (the other dimension). It’s a standard for many Performance Management modules. Alternates exist to compare performance over time, too.

Nancy mentioned that they use bots to collect data and present it in other 9-Box formats. She mentioned how they can:

  • See if the firm is over-invested or under-invested in certain functions (i.e., via experience or salary-wise) and compare that to the strategic value that a function creates for the firm. Would you like to know whether your firm has a lot of very senior, very expensive people in a low-value-add role?
  • See how people and functions align re: culture, retention risk and other attributes versus how satisfied the employees report they are.

Personalized HR – down to each and every individual

What’s so great is that Automation Anywhere sees employees as individuals not just components within a herd of people. They use these tools/bots to create an individual’s “map of hopes & dreams”. This concept is actually huge as most HR groups have no time to develop a deeply personalized career and training plan for each and every employee.  And, even if they did, it would likely be a once in blue moon event. Bots can quietly work in the background and continuously adjust and update this map as new data points become available.

This is a BIG idea. If the average firm has just under 2 ½ HR staff per 100 employees, there’s almost no way for HR staff to develop personalized training, career development, succession plan and more for each and every employee (on top of the dozens of other items HR personnel are responsible for). To do this well would require the collection and analysis of a lot of data – data that often doesn’t fit in a traditional HR application. This data collection is accomplished with the help of bots. A bit of machine learning can develop the personalized plans once the data is available.

This is important as different people have different career needs that frequently change. Employers aren’t necessarily mind-readers and can’t always know where a person wants to take their career, how they want to navigate it, what skills they’ll need, and, what their timetable for change is.

Tools that make HR real-time AND relevant to every employee are sorely needed in the market. You can’t be serious about engagement, creating great employee experiences and possessing a top employment brand if you treat all employees the same and in a rarely changing manner. It’s time to redefine what real-time means in HR.

How HR builds bots

Automation Anywhere sent 30 of its HR professionals to their own school, Automation Anywhere University, to learn how to build bots. Some participants built personal bots and some built more business-like, enterprise bots. According to a recent blog of theirs:

Other bots have been created for more complex duties. According to the creator of an HR application bot, “It’s a compensation- and internal-equity-checking bot that recruiters can use prior to making a job offer. “Compa-Bot” (that is its name) can help ensure pay equity across the company. Currently, this is done manually by a recruiter and HR business partner, requiring back and forth communication. Ultimately, “Compa-Bot” will save a lot of time and increase accuracy and compliance.”

This is the first I’ve heard of an HR organization getting schooled on advanced bot technology and implementing same.  It’s time for other HR leaders to seriously consider doing likewise.

Check out this piece written by Nancy on the use of advanced technologies in HR.

My take

I’ve been frustrated by how long it takes a major HR or ERP vendor to develop and make commercially available a new analytic offering. Most take 18 months or longer for each new capability. These long timeframes make me seriously question the “power” they allege their platforms possess. If the vendor needs more than a year and half for one simple data point, what chance does a HR professional have to do likewise?

The quantity, functionality and development speed of the bots that Automation Anywhere developed are impressive and should trigger HR leaders elsewhere to either create many of their own or to put pressure on their HRMS solution provider to rapidly create equivalent solutions NOW!

Workday and Automation Anywhere are pretty tight. From a February 2019 press announcement, we read:

Automation Anywhere®, a global leader in Robotic Process Automation (RPA), announced that it has received a strategic investment from Workday Ventures, solidifying a partnership that will help connect Workday customers to Automation Anywhere’s world-leading, intelligent RPA platform. In addition to the investment, Automation Anywhere has joined the Workday Software Partner Program.

Let’s hope that these two firms incorporate more of these bots into Workday’s HRMS solutions soon.

Finally, much of the focus on advanced technologies in HR has been around the use (and misuse) of machine learning, artificial intelligence and facial recognition technologies. The bot focus has mostly focused on a few niche use cases (eg: having a bot help a jobseeker schedule an interview or help an employee get an answer to a benefits question). I suspect that we’ve all been too focused on cutting/bleeding edge technologies that may not be ready for prime time right now or may present too many litigation claims if deployed now. Instead, we should re-look at bots and their ability to quickly add value now.

A grey colored placeholder image