The Big Promise of Citizen Developers in Democratizing Automation

Nischay: Welcome everyone, to another exciting episode of the Zinnov Podcast – Hyper Intelligent Automation series – the one-stop shop to hear from the leading innovators and trailblazers within the automation space. I am Nischay, Principal and head of automation practice at Zinnov, and I will be your host for this episode.

The automation industry is poised for disruption and the space is abuzz with rapidly intensifying competition, massive funding, IPOs, acquisitions, and so much more. To help us decode this action-packed automation space and unlock the growth levers to drive greater value from automation, I have with me today, Stephen Siciliano, Partner Director for Power Automate at Microsoft. In his multi-faceted role within Microsoft and his stint across the industry in different capacities, Stephen has emerged as a prominent voice within the automation space.

Stephen, welcome to this edition of the Zinnov Podcast. We are truly delighted to have you with us today.

Stephen: Thanks for having me today.

Nischay: So, let’s dive right in, Stephen!

We know that your association with Microsoft goes beyond the acquisition of MetricsHub, which you co-founded along with Charles back in 2013. It will be great if you can throw some light on your journey so far, and how this pivot towards automation and low code come about.

Stephen: Sure thing. I actually joined Microsoft right out of university. I was an intern back in 2008 and then joined as a full-time employee in 2009. And what I worked on for the first few years at Microsoft was actually – Office. So, I worked on Office for Apple platforms, Office for Mac, Office for iOS. And in that, time touched a whole bunch of different features, not necessarily directly related to the Automation space, but sometimes tangentially related to Automation.

But for a while, I worked on something that’s called ‘OLE,’ – Object, Linking, and Embedding. But it’s really about how you can share content between different applications inside of Office. It’s a technology that’s been around for many years. Of course, inside of Office, you have technologies like Macros inside Excel. I was very familiar with how those work as well. But really worked on all different aspects of office. The things like co-authoring the introduction of the ribbon, the move to subscriptions, Office 365, and that kind of cloud delivery model was a big investment for us at the time.

So really touched on a bunch of different aspects of the Office ecosystem prior to leaving Microsoft and creating MetricsHub, which is really focused specifically on Automation to help people optimize their Cloud infrastructure. And so, went from a broad role to very specific role. And with MetricsHub, we really had a great opportunity to go deep into that area and to identify the needs that customers of Azure had to improve their infrastructure. Things like auto-scaling didn’t really exist inside Azure at the time. So, at MetricsHub, we built the infrastructure to do that on top of public API APIs that Azure had, but hadn’t put the pieces together to create an actual Automation platform on top. That’s why it was very attractive to Microsoft to bring that in as a part of the broader Azure offerings. And as a part of Azure, we were able to extend that Automation to not only be focused on auto scaling, but also alerting other types of automated behaviors inside the platform. And that ultimately turned into Azure resource manager, which is the whole deployment engine that drives, creating resources, deploying updating resources inside of Azure.

Then from there, what we actually did is we said, hey, even beyond the management of Azure, you could actually use this as a generic workflow capability. So, we actually took it from Azure resource manager and created Azure Logic Apps. And from Azure Logic Apps, that was really focused on platform-as-a-service, as a way for people to build the B2B system integrations and Automation in that space.

And then ultimately in 2016, we took it from Logic Apps and said, hey, this can be broader. This could be a SaaS, that could be cloud native connecting to any different service, and that’s when we created Microsoft Flow and is now Power Automate.

Nischay: Wow. That sure has been an interesting journey so far, Stephen. I really like how you described your journey, starting with tangential Automation within the Office ecosystem, such as with Excel, macros and so on.

And today, of course, you are the custodians of the Power Automate platform, which is making leaps and bounds within the Automation space. And digging deeper into the Automation space itself, Stephen, I think Automation has transformed rapidly over the past few years with a lot of complementary technology areas coming together in a cohesive fashion.

At Zinnov, as you’re aware, we have Hyper Intelligent Automation or HIA philosophy, which includes process mining, document digitization, RPA, and Low Code/No Code. Given this construct, we see that the enterprise customers are struggling to make sense of these technology areas. And we know that you have a great coverage across the HIA framework yourself.

How are you advising your customers today on these technology-related decisions as to when do I choose RPA? Or when do I choose API? Or when do I go with Low Code/No Code? Can you help demystify this for our audience?

Stephen: Yeah, absolutely. The first thing I’ll say is this Intelligent Automation space is very interesting because it brings together complementary technologies.

So, there isn’t necessarily always going to be a right answer and a wrong answer in terms of what exact technology you’re going to use; specially between you have the classic RPA, which is that UI Automation. And then you also have, what Forrester calls DPA, which is the Digital Process Automation. That’s really that connector-based technology.

When you use a UI Automation versus talking to APIs, that’s primarily driven by, are there APIs available? Because at the end of the day, RPA is something that is inherently fragile. Using RPA to automate an application, can only be as stable as the user interface of that application. If on the other hand, you have APIs, like we do through our connector ecosystem, that makes it possible to have stable, reliable ways to connect to that application and its data. So, we would first look at, does an application have APIs? If so, then you should be using that set of technology instead of RPA.

But you also mentioned low code. And generally, when we think about Low Code, we’re thinking about applications. And that is a different layer on top. It’s still part of the broader, Intelligent Automation portfolio. But what Intelligent Automation has in the context of Low Code is really, I want to build custom experiences on top of the data that’s in those systems that I need to connect to.

So that’s where, as Microsoft, we have something like Power Apps, which brings in this rich, customizable interface to build experiences that can use the data that’s in those legacy systems. You can connect to it either via RPA, if there’s no APIs, or DPA if there is some sort of API, and then you can create that rich experience for somebody to connect to. That’s really how we see those different pieces coming together.

Nischay: Got it. This helps understand the nuances of RPA or UI Automation, API integration, and Low Code/No Code in a succinct fashion. While this is one dimension to look at this labyrinth, the other dimension pertains to the developers of Automation. That is whether Automation should fall under the purview of the professional developers or the citizen developers. There are two schools of thoughts that we’ve come across. One, which firmly says that automation is meant only for developers and citizen developers cannot build apps without know-how of coding. The other theory, of course, postulates, that Automation has to be focused on citizen developers, in order to scale Automation. Can you help us decode this for our audience a bit? How should Automation – be it RPA or Lw Code/No Code – be viewed as a toolkit for developers or citizen developers or for both?

Stephen: I’m going to take the easy way out and say both. And I can definitively say for anybody out there saying that based on the technology that exists today, somebody who is a citizen developer, couldn’t be successful building out Automation that is not true. With Power Automate, we have millions of people, the vast majority of whom, are non-technical developers. There are people who did not get a degree in computer science. They did not have industry experience building a writing programs. But they’re building Automation every day as a part of improving their own job. Now, that being said, the answer is both because we also do see a huge number of professional developers using Power Automate because it’s so much faster and easier to use a Low Code platform than it would be to open up Visual Studio or Java or some other type of tool, write the code by hand, test that, deploy it, manage the packages, all of that stuff. The beauty of a true Low Code platform, what we have with Power Automate and Power Apps with RPA and DPA is that it targets a full spectrum – all the way from people who aren’t very technical up to the most advanced developers. And we see success across that entire spectrum with the Power platform. The point about, hey, you need citizen developers to be able to scale, that is absolutely true. Today, with what we have it’s simply is not true that citizen developers can’t be successful. That may have been true in the past, prior to the Power platform or the technology that has been out there for a long time to help Automation, but certainly has been more difficult to work with. But I’d say, in the past five years, that’s definitely no longer the case.

Nischay: Got it. Stephen, basically you are focusing on the entire spectrum of developers – both pro-code developers and Low Code developers. Digging deeper into this – do you think this also depends on the complexity of use cases which are being automated? For instance, the high complexity use cases such as within say the IT Ops or ITSM space can be handled by the IT teams or the more pro code developers, and the low complexity ones, such as simple workflow automations can be managed by citizen developers?

Stephen: Yeah, that’s definitely a good way to think about it – is the complexity of the integration requires scaling of the skillset of the people building out that Automation. But I wouldn’t necessarily say that IT is only building complex integrations or that citizen developers are only building simple integrations. But in terms of where you see the major investment, that’s absolutely true. Citizen developers are generally going to be building more lightweight things than IT. But we do see some cases where IT just needs something quick and simple, and they can get that done in a few minutes, using Power Automate. And conversely, there are some cases where we see citizen developers who really spend the time to learn and understand the platform, and they build something that can be very advanced. But you’re right; for the vast majority of use cases, it’s going to be the way you described it.

Nischay: Okay. And Stephen, if you zoom into the citizen developers more, the one challenge that we hear from our enterprise customers is when thousands of employees start building their own automations, then the governance and management of the Automation initiatives, typically goes for a toss. How do you think enterprises can tackle this technical debt, if you will, which is added by citizen developers?

Stephen: This is why having a centralized governance capability that is cloud-native is really important. With Power Automate every Automation that is built by citizen developers inside an organization is stored in a way that the administrators can see what the Automation is. They can see what it’s connected to. They can see if it’s working or if it’s not working. That means that they, as the people in charge of governance, concept policies that control what is okay, what is not okay for the organization. So instead of having to manage every single Automation individually, they can set the guardrails in place that define how the organization should be successful.

Some organizations who have this mentality of every single Automation that is built in my organization, I need to look at and approve. And the challenge I would put on that person is to think about Excel. Every single Excel workbook with a macro generally is not going to be approved by IT. The reason that is a mentality is because with Excel, it’s scoped, it’s safe. It’s only working with data inside Excel.

Which is true, that Automation is much more powerful. But with the capabilities like data loss prevention that we have inside the Power platform, you can make sure that the right policies are in place to make building Automation as safe as it would be, to just have your citizen developers build out an Excel workbook. I would not necessarily recommend that central IT tries to, or even has a goal to monitor and configure and control every single piece of Automation built in an organization. And we see organizations that have hundreds of thousands of Automations running inside of that organization. It’d be impossible for IT to scale. But the ability to set policies and then the ability to monitor and set aggregate views on top of that. So being able to see – is there Automation, that’s running a lot, that has a bunch of people working on it together, that’s been shared very broadly, that uses high sensitivity data sources – you can monitor all of those aspects. And then for just those business-critical processes, step in, get more details and get finer control.

Nischay: Great. So, you need a centralized governance mechanism to monitor and manage Automations as required, while ensuring that IT does not become the gatekeeper for Automation, thereby still fostering this innovation culture.

Shifting gears now, Stephen and touching upon the intense competition we are witnessing in the Automation space. There are clearly two categories of platforms that stand out today. On one hand, we have the incumbent platforms, including the big three vendors players like Kofax, Nintex, WorkFusion and so on. And on the other hand, we have all the technology majors, which includes you, IBM, SAP, ServiceNow, Google, and so on. Is there almost like a market polarization happening with the incumbent vendors on one hand, and the tech majors on the other hand? Or are you targeting different areas? And again, how are you helping your customers navigate this labyrinth?

Stephen: I think at the end of the day, customers don’t really care too much whether or not their Automation provider is part of a broader organization, or is a standalone Automation provider. Really, what matters is, the same things that would matter for any purchase – is this something that meets their needs? What is the price of it? And is it something that will be sustainable and have high ROI in the long run?

I think that when organizations look at their needs, they’re going to find that a company like Microsoft can provide really strong ROI because not only does it have the Automation capabilities, but those Automation capabilities are connected to this broader Power platform, this broader Low Code portfolio. So, you actually see a really rich ecosystem beyond just a strict, Intelligent Automation pieces. And I think that’s something that, that customers recognize as providing a lot more value for the money that they’re willing to pay. But I don’t see it as there’s going to be two separate Automation markets. I really do think that there’s going to be one. And it’s just a question of what vendors can provide the best value for their customers.

Nischay: I like this three-pronged framework that you mentioned, Steven around enterprise needs, pricing, and value or ROI, which an enterprise is leveraging today in order to make platform decisions. I would in fact go ahead and add another dimension to this, which is around system integrator partners or GSIs, which are actually implementing Automation for these enterprises. And hence, oftentimes end up influencing them on their platform investments as well. We understand that you as Microsoft have been a purely partner or GSI-led organization. In fact, you rely majorly on partners for your go-to-market motions. However, when we look at the Automation scenario, I think you came into the scene only last year, and your existing partners had already made significant investments and commitments with some of the other competitors, especially around joint Center of Excellence setups, and certifications, and so on.

Given this context, how are your partners or GSIs seeing Microsoft within the Automation space? And are you putting massive investments in existing partners? Or looking to go after net new set of niche partners for Automation?

Stephen: Yeah. So as Microsoft, we can never afford to be draconian in terms of requiring that everybody goes our way or the highway, so to speak. We’re a big company. We have lots of partnerships with a lot of the other RPA vendors. So, it’s not necessarily an either – or. We believe that all of the partners that are out there in the Automation space, they’re going to find strong opportunities adding Microsoft to their existing portfolio; not by trying to remove or shut down their existing practices and switching entirely to one vendor like Microsoft.

So, it’s going to be an open ecosystem. This is reflected even in the way that we as Microsoft, have connectors in our ecosystem to all of our competitors. So, I don’t think it’s going to be an exclusive thing. All of the partners are going to really pick up all of the different options that are out there.

Nischay: Got it. So open ecosystem and co-opetition seems to be the mantra for the Automation space, it seems. But another area which has a rapidly gaining prominence within the Automation space is around process discovery. How do you select the next best processes for Automation at scale? And while there are a bunch of these scientific tools, such as process mining and task mining, which are available today, we understand that enterprise customers still continue to rely on the more tribal or manual techniques, such as outsourcing, conducting workshops, hackathons, and so on to identify processes for Automation. What has been your observations in this context? And we understand you announced your Process Advisor tool recently. How are you solving for such manual techniques, which are clearly being preferred by your customers today?

Stephen: I’d say there’s two reasons why an organization might prefer the manual approach. One is that it’s a lower barrier to entry. In some ways, it’s easier just to ask people what they want than it is to deploy something and do a big lift in terms of data. The other one is, oftentimes, companies will have high expectations in terms of privacy and trying to implicitly monitor and figure it out what it is that their employees are doing, makes them have to ask some tough questions in terms of what is it that their employees would be comfortable with. I think that’s why we see a big effort for collecting ideas in a manual way. So, what we have actually with Power Automate, is the Center of Excellence toolkit. And one of the capabilities that toolkit has is an Ideas Hub, that allows this exact type of manual motion of people just going, submitting ideas, voting, getting the different crowdsource mentality. So, I do think that is one important track because of the potential concerns that organizations have.

But at the same time, we have Process Advisor in preview right now. And what Process Advisor does is it gives the control to the regular business user to describe the process that they’re doing, and then submit that to the central repository. And by putting that control in the hands of the citizen, it means that they are confident that whatever it is that they’re doing is monitored in the way that they expect. And then, the central group can take a look at those different recordings and understand how to potentially automate them. So, I do think both approaches are going to be important, and they’re not even mutually exclusive. You can submit ideas that have recordings associated with them. So that way you can provide additional details on what it is that you’re trying to do.

Nischay: Great insights, Stephen. I think what you described as a hybrid approach to process discovery makes a lot of sense – how you crowdsource ideas as well as deploy process mining tools, while keeping in mind the privacy and comfort of citizen developers. This also ties in closely with change management, which of course is a critical lever for successful Automation.

Stephen: Absolutely. Making sure that the business users have the ability to define their own destiny, empowers them to be much more involved in the way that the organization rolls out change.

Nischay: Got it. And final question, Stephen, before we wind down. What is the roadmap plan for Power platform over the next six months? And are there any technological innovations that you are personally excited about?

Stephen: There’s a bunch of things that we’re working on, of course. For Power Automate, some of the biggest things that we’re focused on is getting Process Advisor to general availability. Also, we’re going to be adding new capabilities to that and making that available for broader scenarios.

Another one is deeper integration with the rest of the Microsoft ecosystem. In particular, you’ll see Microsoft Teams, for example, and the types of things that we can do in the context of that ecosystem. Because in some cases, the pandemic is getting better, and in other places, there’s lot of churn there. But we think that people will continue working remotely for quite some time.

So, Microsoft Teams and the ecosystem there, about making it easier to build apps, to build Automation, to have them run in the context of Teams is a major investment. The other area that we’re working in Power Automate is making it just easier and more straightforward to build out Automation.

Making sure that there’s all of the building blocks. That way citizen developers can be more successful. Especially now that Power Automate desktop is included with Windows – that’s going to be a very broad audience of people who have a lot of expectations in terms of ease of use and simplicity of getting started. Oh, and there’s one more area that I would say is a major investment, which is really the governance, and looking at an enterprise-wide view, making it easier to deploy capacity for the Power platform, just giving more advanced analytics and reports. Those are another, a major area that we’re working on.

Nischay: Thanks, Stephen. We definitely look forward to some of these announcements from Microsoft.

Well, that was an extremely insightful conversation, Stephen. I think your viewpoints on the confluence of technologies within Automation, the democratization of Automation with citizen developers, your insights on process discovery, change management, and how you see the space progressing, but really interesting. And I’m sure our listeners will have some great takeaways from the session. Thanks, Stephen, once again for joining us. It was a real pleasure hosting you today.

Stephen: Great. Thanks so much for having me.

Nischay: With that, we come to the end of this podcast episode. Thank you all for tuning in and watch out for more such engaging discussions in the Zinnov Podcast series. Have a great day, everyone.