What is enterprise automation? Q&A with Workato’s Markus Zirn
The enterprise landscape is poised for a transformation thanks to recent innovations in a technology concept with roots that go way back: automation. In this post, Storm Ventures managing director Ryan Floyd explores the power and potential of automation with Markus Zirn, head of strategy and business development at Workato, a Storm portfolio company that recently completed a funding round.
Let’s start with a basic question: what is this idea of automation? We’ve been hearing a lot lately about RPA and enterprise automation. Can you break it down for us and put it in context — both historically and where we are today with the technology?
Absolutely. My involvement in automation and business processes actually dates way back. My first job out of college was in something that was called BPR at the time: business process redesign. I was working for Booz Allen Hamilton, running around enterprises and modeling their processes. I would interview people to understand how they did things in the company and then create Visio diagrams documenting that particular business process. The idea of automation is to model a business process and take away as much manual work as possible. Ideally, the process would be something fully digitized: something straightforward where a customer puts some data into a web form and then it moves through the system. That’s actually the best possible experience for a customer, where everything just gets done automatically. A good automation platform helps you with this purpose.
There have been entire generations of automation tools. If you go back to the 90s when I was doing BPR work, you modeled processes and then there was someone else: a developer who would implement those processes using code or very old integration platforms or ETL (extract/transform/load) tools. You mentioned RPA; that’s been a new trend lately. You could even call it an automation hack if you want.
What does RPA stand for? How does it fit into the context of what people thought of as automation back when you started compared to today?
RPA stands for Robotic Process Automation. It’s a new way to automate. It’s kind of a simpler way to do it. The idea is that you simulate a person’s clicks on a screen. So a robot is basically mimicking a person, executing all the keystrokes and clicks that happen almost like a synthetic user.
What’s interesting from an IT perspective is that it makes multiple things possible. First and foremost, which a lot of people like: it’s easy to do.
Right, so you don’t have to get a developer to implement it.
Exactly. You just record those screen movements. It actually cuts out all the back office IT. All the APIs, all the infrastructure — all of that is cut out. But it has some disadvantages, too. For one thing, it’s hard to scale. It’s harder to do more complicated processes. It’s more for smaller tasks; it really is down to the speed of a person.
(For anyone who might be new to the concept, we should clarify that it’s not actually a robot doing the work. It’s a script; we call it robotic process automation, but it’s really a script that’s making the data entry much, much easier to do.)
You’ve been in automation for a long time, and there’s been a number of companies that have built significant businesses around what they would call automation over the years. Sometimes I think it seems like these technical concepts come out of thin air and just hit the scene, whereas in reality, there’s more of an evolution where things grow over time and infrastructure like the internet and the cloud make new advances possible. Can you talk a little about automation in that context? What were some of those earlier companies that did automation, and what did it mean to be in that space ten years ago?
The late 80s were when integration tools started off: something like TIBCO or Informatica. The idea was always to remove manual steps that move data from one system to another. What we didn’t have back then was the APIs that exist today.
What’s an API?
API stands for Application Programming Interface. You can think of it as a mechanism built into an application to make it more easily integratable with another application. Anyone can use a software application as a person by going to the screen and doing things. The idea of having an API is to provide another door into the application where you don’t have to go to the screen, but you can still do the kinds of things that you would do as a person. Except instead of a person, you can write a piece of software that will do those things for you. So you can think of an API as a remote control for an application that helps you do the things you would normally do, but at the speed of software.
It’s kind of like Legos: you can build APIs for applications and get them to talk to one another easily.
That’s a good way to think about it. If you’re trying to automate a business process that has multiple applications, and each application has APIs, they’re like the knobs on the Lego blocks: they let you fit the multiple applications together.
What’s a simple example of automation from ten years ago?
Let’s talk about business process examples. The fundamental business processes haven’t changed. One of most common ones is order-to-cash, which starts the moment a customer orders something and traces all the way through all of your systems until you get the money and the customer’s product gets shipped. Another good one would be any kind of onboarding process, where a new employee joins the company and gets entered into all the different systems, or where a new customer gets onboarded.
All of these processes are still the same today as they were yesterday, so you always had to integrate. For order-to-cash, for example, you had to integrate your customer relationship management (CRM) system with your enterprise resource management (ERP) system. For onboarding, you’d need to integrate your HR system with your identity system or your IT help desk system. So those processes haven’t changed. How intelligently you can do it today, that’s changed quite a bit.
So these enterprises have all these different systems that control critical functions of their business: how they take orders, how they record those orders, how they manage their inventory. And you need some way to connect them, which you could do manually, but it’s very labor-intensive and error-prone. This is where automation comes in, because it connects all these pieces together better and faster.
That’s exactly right. Automation is about removing manual steps. A manual step is always a point where a mistake can be made, and it takes time. You might have to wait for someone to take that manual step, and it might not even happen today. All of this affects the customer experience.
Software doesn’t care if it’s two o’clock in the morning. It’s happy to enter the logic at any time of day very quickly.
That’s a good point.
Let’s pull forward to today. You’re at a really interesting company: Workato. What’s different now about what Workato does compared to these companies from several years ago?
In starting Workato, the idea was to create a better platform: a new platform that was much more powerful, not just in an enterprise sense, but also in an ease-of-use sense. We all brought different experiences to the table. I modeled business processes in the past; Vijay, our CEO, co-founded one of the first integration platform companies. We felt that the idea of automation and integration was still very similar to what we knew back then, but the environment around us — all these business systems like CRM, ERP, and so on — had actually changed dramatically. Salesforce was the first to say “No more software; let’s do this all in the cloud. Let’s make it much, much easier.” Salesforce was much easier to use than Siebel, for example. In the HR space, Workday created a much better experience than PeopleSoft. NetSuite brought financial systems into the cloud. And while all of these processes were all hidden in their own little silos in their own little departments, they still had what we call operational processes or operational workflows. Everything that the customer actually notices, experiences that cross multiple applications, only had old tools to do that. So our goal was to say: how can we lift that into the modern world? How can we leverage, for example, the fact that everyone now has APIs? If all these service providers are in the cloud, why does your automation tool have to be on-premise? That didn’t make any sense. So we modernized what we knew and created the platform that we felt was a path to the future to really digitize these companies effectively.
One really simple example that I’ve used in the past to explain the idea is smart outlets. On your phone, you can use an app to set up triggers that can turn your lights on and off when you want to — and I don’t need to hire anyone to program my phone to do that. They make it really easy for me to do that myself. So now I can set lights to go on and off when I go on vacation, or control their color temperature with all kinds of basic logic that’s pretty simple to add. Obviously, Workato’s much more complicated than that, but at the same time, it doesn’t have an interface the same way Salesforce does. I don’t think it’s quite as easy for many people to get their heads around what it looks like; can you give us a sense of how Workato makes things happen in modern SaaS applications?
Workato has simplified and improved the automation of business processes. Your smart outlet analogy’s vacation use case puts to work something we call a trigger. When a lot of people first look at Workato, they say it looks like If This Then That for enterprise, and that comes down to the base unit of Workato, which we call a Workato recipe. A Workato recipe has a trigger that defines when that recipe, that automation construct, should kick in. Workato listens: it listens to data, to events in applications, to exceptions. And when it hears something it’s listening for, then it kicks off what we call actions. That’s the other side of the recipe: trigger and actions. Actions are your business workflow with all the steps, branches, loops, and everything else. But Workato can actually access all your applications: your SAP finance system, your Salesforce system, your data warehouse. The base unit, the recipe, acts as a kind of Lego block of automation. It sounds simple, but if you put many of these recipes together, you can get amazing models of complex business processes, all working together.
Just to give you an idea of the scale at which Workato functions, we just got certified by Snowflake to the point where we can actually execute half a million starts of these recipes. That’s half a million trigger events every single second with Workato, for a given customer. So this isn’t just for a light bulb turning on and off; Workato can do big, big things. We can connect to Salesforce really easily. Some people might think that’s not that hard: it’s a SaaS application, it’s got a beautiful API. But we can do the very same thing if you have an SAP system from the 90s that’s still running behind your firewall; we can access that just as easily. Workato’s built as an enterprise tool no matter where your enterprise runs.
A lot of what you just mentioned goes a long way to realizing the concept of the intelligent enterprise and an enterprise OS. Most everyone reading this is bound to be at least somewhat familiar with operating systems that run on our local computers like macOS or Windows; tell me how you think about the enterprise OS in the Workato context.
The enterprise OS really came from the idea of contrasting enterprise IT and your personal IT, which is your laptop. If you think of your laptop, you have applications — Word, Excel, PowerPoint, and so on, just like the enterprise has the CRM, the ERP, the HR system. But your laptop has an operating system like Windows or Linux or macOS, which has all the tools needed to make all those applications work together. For example, you can copy and paste between different applications on your laptop; the operating system provides that. And ever since the 80s, the operating system has been getting easier to use, to the point where just about every one of us can use it: you have your mouse and you drag and drop stuff.
Today, in the enterprise, it’s still not that way. First of all, you don’t have one OS: you have different, separate tools. You have an integration platform, you have an ETL platform, you have a business process management platform, you have an RPA platform, and so on. It’s hard, and you still need developers to make them all work together.
Our goal is to build an enterprise OS like the one you have on your laptop: something as easy to use as the macOS that combines all the pieces you need so you can build great customer and employee experiences across all your business systems.
Can you talk about your favorite examples of how customers are using Workato?
One of our customers is the government of New South Wales, the largest state in Australia. Something probably every one of us can relate to: in the past, when you applied for a government service, it was 100% manual. You had to write letters. You had to call. Eventually, there was a website where people could submit a form. So what Workato did was tie that form to all the systems in the background that the government was running: there was a content management system where all the submitted information was stored, a Salesforce system where all the citizen data was kept, and a financial system that could actually award grants and provide subsidies or whatever else the people needed. Workato converted that government process from 100% manual to 100% digital.
HubSpot is another good example. They provided customer success, and they had a Salesforce system for that. They had a lot of information in Salesforce, but the customer success manager still needed to look into the system to identify situations where a customer needed help; finding the right moment to jump in and interact was a manual step. That’s what Workato changed: we built Workato recipes that connected to a Snowflake data warehouse with all the customer information, and whenever something deviated from a great customer journey, a Workato recipe would wake up, notify a customer success manager, and automatically set all the right fields in the support app, the marketing app, the CRM app. It made them more data-driven. It gave them superpowers; before Workato, they needed a lot more customer success managers to handle the same work.
If you believe that the enterprise is going to have lots of applications going forward and there isn’t going to be one single vendor to rule everybody, then there really needs to be some way to be able to connect all these systems together. It seems like something like Workato would almost have to be the future for virtually every single enterprise.
Very much so. We see companies with 700–800 employees that have something like 250 SaaS apps. It sounds somewhat ridiculous, but it’s the reality of today. The economics of SaaS really drive SaaS applications that are best-of-breed. You’re rewarded as a SaaS company if you build something that solves a certain, specific need really, really well and you go deep; I think that’s when SaaS vendors become successful. And as a result, we see more and more SaaS applications out there.
We even see it when older companies transform their business. MGM became a Workato customer recently. They had SAP and were writing a lot of custom applications. They still have SAP for financials, but now they have Salesforce for CRM and support. But they also have more media-specific vertical applications, like royalty management, media asset management, and so on. All of these systems need to be held together, and frankly, you want these applications to talk to each other. Just like HubSpot, you want customer data that might not even be in an application and just lives in a data warehouse. You want to be able to consider that as part of your operational workflows.
That’s the dirty little secret. The operational flow that really impacts the customer experience, like order-to-cash or a support flow, rarely interfaces with just one system. It always includes multiple systems and data from multiple applications. So you need to have something to face that reality.
Everybody out there has heard about how important data is today. We’re gathering so much data. But in order to really take advantage of the data that you’re harvesting and creating, you have to be able to exchange it between all the different systems you’re running. If you can’t, then it just sits in a data warehouse, and it doesn’t provide a lot of value there.
Absolutely. There’s so much powerful data out there. A lot of SaaS companies have a lot of product usage data that tells them what people actually do with their product. It’s extremely enlightening, but if it just sits there and all you do is look at it, it’s only worth so much. An automation platform like Workato can listen for certain patterns, take a specific customer set, and load it into a tailored marketing campaign based on their product usage. That’s very powerful. But you need a tool like Workato to make it happen.
With 200+ SaaS applications, you get 200+ disparate silos. So you’ve got to bring them together; Workato can do that. Now you have to listen to all that data for some specific trigger; Workato can do that, too. People often bring their data into a data warehouse and then they put Tableau or some kind of business intelligence tool on top and think “Okay, we got our data all in one place.” That’s great, but they still haven’t operationalized the data. They still haven’t said “If x happens, go to your CRM system and your HR system and put y data in.” Getting to these operational, data-driven workflows is a huge opportunity.
There have been all kinds of attempts to solve the problem, like with integration platforms, but often they’re too technical. Or solutions like RPA, which seems good but doesn’t scale. We really believe that with this unit of the recipe, we can make a real dent in how companies can easily improve their operational workflows and create an intelligent enterprise.
Written by Ryan Floyd, co-founding MD of Storm Ventures, which invests in B2B tech across the world, and is the host of the #AskAVC podcast and video series.