A champion for change: Spotify’s global CoE leader talks hyperautomation with WonderBotz

A conversation between Sidney Madison Prescott, Global Head of Automation at Spotify, and Paula P. Carneiro, Director of Partner Relations and Business Development at WonderBotz, on Hyperautomation Center of Excellence leadership.

You have had a broad range of automation-related leadership roles throughout your career. Tell us a little bit about what led to your current role as the Global RPA & Hyperautomation Center of Excellence leader for Spotify?

I have been within the Intelligent Automation field for almost six years now. I started out at First Data, within the global software and engineering organization.

From there I moved to a data quality and governance program manager, which led me to being invited to stand up the company’s first Global Intelligence Automation Center of Excellence. That began my foray into Intelligent Automation.

And with that program, I was specifically dealing with RPA (robotic process automation). We also had chatbots. We were working with IBM Watson and were established in the UK as well as in the U. S. It was the beginning of understanding the global enterprise implications of an automation program.

From there, I went over to BNY Mellon, Bank of New York Mellon for those who aren’t familiar. I was a part of our Global Intelligent Automation Center of Excellence there, with a population of more than 400 bots in production, which set me up to take on Spotify in early 2020.

We often advise our clients who are just starting out to look at Finance and Accounting for initial use cases and to build momentum.  What does RPA in Finance mean for you?

RPA, in my experience, is really about financial engineering. I’ve worked predominantly with the stakeholders that sit within finance. Everyone from Tax to Treasury to Internal Audit, and to FP&A. To those teams, I bring my area of expertise, which has been foundationally around RPA.

Add to that my experience with OCR and chatbots, machine learning engines for the OCR platforms, and it allows me to be knowledgeable beyond core RPA with tools that can truly transform the way Finance and Accounting is done.

So, since it seems to be widely agreed upon that F&A is a good place to start, how do you bring reluctant senior leaders along with you on your RPA journey?

Over the years, I’ve seen my role as a partner to the technology. However, a lot of times, my CFO clients are the first use cases as you mentioned before. Typically, CFOs get involved when the organization has already picked out an RPA platform. The Digital Transformation drivers have already figured out their methodology, stakeholders and doers, and know how they want to approach it.

The finance function is key because they have the most manual, repetitive tests with tasks which, as you know, make for great automation candidates. Oftentimes, they’re called upon to do a process review, perhaps explain how something is currently done in its current manual state. Your first role is to say, “Okay, that’s automatable” or not.

What is your advice to somebody who’s just dipping their toes into understanding RPA? What should they look for to get started with a program?

When you’re starting out the foundation of a Hyperautomation Center of Excellence with RPA, the very first question that I always ask myself is, “What is the measure of success for this program?”

Then, the second question is “And, why are we doing this?”

More importantly, we want to go on a journey of digital transformation. The “why” is extremely important. That’s really going to guide the roadmap you create for your stakeholders and the different business groups within your firm.

Points like “Where do you hope to be in six months? Where do you hope to be in a year? Where do you hope to be in five years? And even more importantly, how are you going to manage the change?

It’s an institutional change to really bring in these types of automation, specifically in the financial sector, where we’re so heavily regulated. And it’s a massive effort that you need to make sure that you’re in alignment with your internal audit team, and with risk and compliance, and that you understand all of the global policies regarding data, and the way that you’re actually going to manage your bots.

So, there’s a lot of legwork that you have to do to understand things.

Number one, where are you today? The second one is, where do you hope to be? And the third point to really consider is, how am I going to get there?

RPA, well, you can consider it as one tool in your toolbox.

As we start diving into Intelligent Automation, you’re really going to be looking at your optical character recognition (OCR) and natural language processing (NLP). You’re really going to be looking at machine learning (ML) engines that can optimize the way that you can take documents that are not machine readable and transform that format.

When you first start out, it’s really important to just say, “What am I hoping to achieve? Where are we today? And what is the methodology? What are the steps that I’m going to take one by one, quarter by quarter, year over year, to actually start progressing towards that final transformation of my organization?”

How do you keep the CFO engaged in RPA/Automation initiatives?

I was on a call earlier and this point was actually brought up. CFOs are always characterized as trying to find that ROI and cost cutting. So, it is important to remember that as you assess the “why” for each stakeholder.

For a CFO to understand the investment, you must tangibly show them how this is going to scale. How is this going to grow?

A lot of the woes that CFOs have include, for example, marketing is asking for more money. Where can they show us the tangible ROI of where that has been?

You know, there are so many different avenues that, once you enable automation as a mindset in your organization, the sky’s the limit on the data you can extract, to support your growth and ROI.

What was the most eye-opening data moment, like ‘Wait, robots did what?!’ you’ve had from an internal client?

I do have quite a lot of those! I go back to First Data. I can also go to BNY Mellon and eTrade.

There were so many processes that we had that were extremely manual, and within that manual effort, there was the length of time that it took for the end-to-end process to actually complete. That actually had an impact on clients.

We had a lot of different instances where we had SLAs that we were not meeting for one reason or another. The biggest thing that I’ve seen in a positive way in terms of transformation has been the ability to actually change the way that we approach SLAs and thereby change the way that we relate to our clients – both internal and external clients.

I’ve had everything from increases in efficiency from 50% to 75% ‑ really dramatic increases because of the fact that you can put automation in place.

If you are, if you accurately understand all of the business exceptions, you understand all the system exceptions, you really can optimize the process to the point where there’s very little downtime in terms of processing, and collection of data.

It’s quite fascinating to see how often we see such huge gains in efficiency just by putting in the right bot process within the right business department and the right kind of function.

We have seen dramatic rates. I mean six figure rates in terms of the amount of cost savings that we’ve received back as a result of mitigating a lot of additional overhead charges that we were being charged due to missing SLAs. That’s a huge savings to the company.

And that’s only one department.

Wonderbotz Back-office Bundle icon use to add fast value in your RPA Program

I would imagine that if you do that across the firm, then you’re looking at potential savings in the millions just for compensating for compliance dings and inefficiency. Does that resonate?

Oh, absolutely. I think that’s definitely the way to look at it because the risk averse CFO, when we kind of make it more plain text, and just evangelize the fact that there’s so many benefits to eliminating checks and re-checks, for example, it’s back to the reimagining the way that the process is done. That’s where the conversation always lands.

You no longer need a calculator to rectify whether that reconciliation worked out well, or to trace back to what that charge back equalized, because a machine is not going to get it wrong. So, compliance and risk averted are two of the big drivers that we see a lot.

I think a lot of people attribute a robot, as a 1-to-1 replacement of a human, and it’s really not.

It’s a creative, iterative technology that enables the user to just stop and think. The manual, repetitive tasks are getting done so you can stop, analyze that data, and report back to the organization the stuff that actually matters.

Aside from some of these tangential benefits, what really got your Finance Department excited about RPA? How did you evaluate the use cases specific to RPA?

It begins with understanding the complexities of finance and the fact that we do have so many manual, repeatable processes that we have our internal employees working on every single day that do have an impact to, let’s say, the revenue of the entire firm. These are also processes that have inherent risks attached to them.

That is one of the reasons why we spend a lot of time on them, right. We’re checking, we’re rechecking, the data that we’re entering, and we’re making sure that the reports that we’re sending up, the financial reports to our CFO, are accurate. We have entire teams just dedicated to putting together those revenue reports.

But, in our desire to showcase a very clear picture of what the company looks like from a financial standpoint to our CFO, we’re spending an inordinate amount of time working on most processes, and that’s really where RPA can help in the finance sector.

We take all of these processes that we are deliberating over, that we have teams of 50, 60 people working on, just pivoting data and excel and pulling data from different databases. Those are the processes that we automate.

And, typically, I like to start smaller and take some of the work that we may be doing-let’s say offshore-you look at that and say, ‘Oh, wow, this is a completely manual effort. We have an entire team sitting over in wherever they’re sitting and they’re just running through and it’s very rules based.’

A lot of times they have time constraints. They know exactly when they need to get the reporting done, and they’re just walking through repetitive steps of documentation, pivoting data, dumping that data into a database and then pushing that out to another reviewer. Those are the best processes that we can automate because typically, those are low- to medium-complexity processes, which means that on average, it will take 4 to 6 weeks to develop another, you might be looking at another week in UAT and testing, and another week or so in hyper care once you push it into production.

So, you get a really fast turnaround in terms of pushing through the development lifecycle and the benefits are huge. We’re talking about potentially delivering savings where you can actually cut down the number of resources that you have sitting offshore.

Now you can start looking at the same types of daily work that are very manual, repeatable.

If you have an entire team just dedicated to verifying data, that is a perfect process for RPA. And,  within finance, we have so many processes of that nature.

Start small. Start with a particular area where you feel that you clearly understand the business processes and define the rules. You understand the steps, you understand the risk, any risk that’s associated to that particular process. Understand, for example, whether that process is a SOX regulatory process.

What do you do with the time saved?

You start freeing up more time for your associates. Then you can really start diving into how you can benefit from that time savings. You can pick up a wide variety of tasks that the finance organization hasn’t had time to dedicate to as a result of RPA.

I think it’s looking at and understanding what processes are rather mature in their offering and then also just understanding where you want to be.

Then consider, what area do you want to start out in? Do you want to start out in your tax department? Do you want to start out in your treasury department? Are those departments ready for that transformation?

What do you recommend around preparing the workforce for the changes you are making?

It’s all about preparing yourself and paving the way to smoothly transition from the manual process to the automated process and then also preparing the workforce.

Preparing your workforce for the alleviation of those processes, and what else are you going to have them work on? I think a huge lesson that I’ve learned in the finance department is, I always ask, you know, what are the top initiatives and objectives that you wish your team could work on that you can’t today because of your manual work?

And, what I’m doing by asking that question is, I’m trying to empower my finance leaders with an understanding of: Okay, once we place this automation into production, it’s going to open the floodgates in terms of how much time your team has. And, I want you to quickly be able to take advantage of that time.

You have in your mind an understanding of where you’re going to place the team and how you’re going to leverage those resources once you free up a certain percentage of their time at a global scale.

It’s all about priming those finance leaders to understand that once the transformation comes, you’re going to be behind the curve if you wait to figure out what you want your resources to do next, once their time is freed up,

Tell us about why you built the Hyperautomation Center of Excellence (CoE) the way you did and how to begin delivering value to the organization?

If you have a centralized hyperautomation COE where you have your developers who are typically sitting within a specific department. Say, they are sitting in IT, and they are working directly with the business stakeholders across the enterprise, pulling in opportunities, prioritizing those opportunities, and then pushing those opportunities out.

Now, the other way that you can also build the COE is by having a federated model and with that, you can actually have citizen developers who are basically business stakeholders that you’ve upskilled on the RPA platform. Those business stakeholders are overseen by the COE. However, what it allows you to do, is to have a business lead who has all of that amazing knowledge of the business process combined with the knowledge of the RPA platform. Putting those together is where the real power is.

Then you have someone who can develop very quickly because they already understand the ins and outs and the nuances of the business process, and you coach them through and train them up, get them certified on the RPA platform.

From there, it’s a very fast turnaround. You get to a point where you could have new automations rolling out every single week. It’s really just dependent upon the number of resources that you have, who are concentrated on development work, across different organizations within the firm.

Then you literally get to a point where you’re delivering every single week and it’s, you know, it might be finance. It might be HR. It might be legal. It might be operations. And that’s really the ultimate goal.

You’re delivering value in one form or another, and it’s also the way that you can quickly turn it around so fast because the number of resources that you have are developing and making sure that you’re smart about picking your use cases.

Making sure that you understand the complexity and the feasibility of the use case is critical. That can really dent your ability to deliver quickly.

Lastly, as a little “warning” from your experience, what is the sure way to fail as it relates to an automation program?

You will fail by choosing a use-case that’s actually not feasible to automate with RPA, or by choosing a use case that may be too complex for a first-time citizen developer to tackle. You really want to keep those considerations in mind as you prioritize and build the opportunity backlog for all of your business stakeholders. The key to value is a healthy pipeline of automation candidates, so communication is key.