CEO and co-founder of Krista Software, John Michelsen, recently visited with Kevin Craine on The Digital Transformation Podcast. The Digital Transformation Podcast interviews authors, best-selling thought leaders, and top-shelf executives on driving digital success. John and Kevin about:
- What makes automation intelligent
- How to transform processes to enable more thinking and less doing
- Citizen developer projects
- Managing change during digital transformation
Listen to the entire conversation here, or read the transcription to find advice on how you can transform your business.
What do you mean by Intelligent Automation? And what makes it intelligent?
We define Intelligent Automation as an automation platform with intelligence. Tada! But the goal is artificial intelligence. Our mission with an intelligent automation platform is to identify an outcome that currently may take six steps. And those six steps might be because systems reflect the organization and the number of manual steps people are doing. What could AI do to take that six steps down to three or two, or instead of taking two weeks, it’s now taking three days. Intelligent automation is about optimizing the existing business. And that’s why it’s useful for digital transformation.
How do we transform processes to enable our knowledge workers to think more and do less? And what role does Krista play in all of this?
Krista, from its conception, was all about the notion of freeing people to do meaningful work and having the machine do more of the work. And there’s a very fundamental aspect of this.
Intelligent automation embodies the steps, the rules, the processes, the who needs to talk to what, whether that is a machine or a person. That takes the complexity out of your people’s heads and makes your organization run at machine speed, not human speed.
Think of it as right now IT dumps a bunch of tools on a desktop and says you sort out how to run a business. We want Intelligent Automation to run the business and invoke people when they are needed, just like we currently invoke a system when it is needed. But Intelligent Automation is about collaborating with people, systems, and AI. That is the key to unlocking what you just described. Too many times, people are doing redundant labor. Humans get worse at a task the more frequently they do it. Machines, especially AI and ML-driven machines, get better at it the more they do it.
So we want to shift those repetitive tasks clearly, but even outcomes. So when we get to that level, we’re shifting the economics of a business. The machine is doing much more of the work. And people are driving it from a creative, organizing, planning standpoint, but not doing much of that.
Can you give us an example of how shifting work from a person to a machine plays out?
Yeah, absolutely. A CFO of one of the largest banks in the world described digital transformation to me in his view. He stated currently, 80% of his people’s time is spent doing things, and 20% is thinking. His definition of digital transformation is 80% of his people’s time thinking and 20% of their time doing. That’s a fantastic way to capture the whole concept. And that I believe is the mission that Intelligent Automation signs us up for at where we need to go as an industry.
I could give very practical examples as well. We’re working with a company that has 1000s of emails a day coming into various inboxes. And people are, of course, chasing down who would know that, how can I do that?
All the while, they are documenting each email into case management systems. There’s a massive effort associated with dealing with inbound inquiries, disputes, requests to change whatever it is. Now Krista reads those emails. With its ML capabilities, Krista classifies it by type, the topic, and the remediation required to satisfy that request or whatever in that email. Krista logs the inquiries in the case management system, in many cases, resolves that case and marks it closed immediately. This frees up enormous amounts of time, and I can think of just simple things I need. What’s my current customer balance? Can you move my order to Friday? I have to cancel my next appointment. These are things that machines should handle. Why are people chasing down people trying to figure out what systems to use to get those things to occur? So this is a typical pattern in every business. And it’s something that we do quite a bit.
It’s like taking the robot out of the human.
Yes, taking the robot out of the human Yes. And by the way, the robot is just essentially trained by humans. So it’s really funny when we hear these comments around, robots will take over. No, they’re not. It’s no different than having someone on your payroll to whom you were delegating something. The intelligent machines learn from us and then do things as if we were the ones doing them. That’s all upside. Right? I don’t mind at all.
If someone wants to do something three, four, or five times and says, Hey, let me do that for you from now on. I don’t see that as anything but good.
The goal around that is, of course, that I’m now free to do more of the thinking that’s going to make, in my example, my business better, more creative time, more planning time, more awareness of people, and people issues time, all that good stuff.
And it can play out in a competitive advantage. For example, as a customer, I might get my issue solved more quickly and more completely by working with an AI system instead of a person who may or may not be connected and as quick to resolve my issue of giving me the service that I need.
That’s absolutely true. And today, we’re in an evolutionary moment where most automation platforms are quite task-oriented and very heavy in programming to accomplish even those tasks. So, in fact, it’s usually a frustrating experience with a chatbot of a website, right? It’s usually difficult to bounce around a call center trying to get someone to deal with a more interesting question than a very simple one. And as you get closer and closer to where we go with Intelligent Automation, those things can go away. Much more can be done via automation, as opposed to humans. And we end up in a better spot, place all the way around.
Now, John, we’ve had several guests on the show recently discussing the advantages of no code and low-code application development. Can you describe the concerns with citizen developer projects?
Yeah, yeah. So this was a bit of a hot topic for me. And it’s actually a bit of a frustration, frankly. Show me a good low code or no-code platform, and they’ll talk about how they’re a good thing.
Go to any job board and look at the job postings from companies looking for those so-called Low code or no-code platforms. They require five years of job experience in three years of SQL Server. You’ll need to know HTML and CSS. Frankly, the problem with citizen developers is that you’re turning citizens into developers. And also, developers were already citizens. So the phrase has no meaning.
What we actually have to do, Kevin, is to define where there is a place for code. When you’re talking to systems, code is required. There’s no escape. But for the process itself, the rules, the steps, the who knows, who needs to be involved, and in what way and the order of the sequence of events that need to occur it all of that, that is content, that is not programming. So when you separate those two, you have a completely different experience. That is nothing like code experience.
That’s what we describe Krista as. A nothing like code experience for the process automation piece that connects to your CRM, your supply chain systems, and you’re whatever, you’re going to build that as code. You’re going to write it once. You’re never going to have to write it again unless of course, you change the system. But that is different than the rules, the steps, the process. So we do this all the time. We say, Okay, Mr. Customer, you’ve been doing it. There’s been this bespoke lots of code way, right? What if the rule now changes that approval is required at $50,000 right now? What if approvals only required $40,000? Well, I’ll write up the requirement, I’ll put it in the backlog, and weeks and weeks will go by before, hopefully, the development team will get to it. We’ll show them how to do that in Krista in two minutes. It’s not even a testable event from an IT perspective. So that’s the difference. No code. Actually, no code is exactly where we need to be. What we see in most of those platforms, is paste your code here. That’s what we call it.
John, you say automating what you do today is the antithesis of digital transformation. What do you mean by that?
We see a lot of projects that customers undertake, and they call it a transformation, by the way, where they’ll do some level of task discovery or process mining. And then they’ll say, okay, here are the steps. And here’s what happens on each desktop for each system. And then they’ll go about automating some of those tasks, where they’ll have a bot do that instead of Sally or Joe, right? We’re going to be serious here.
That’s not transformation at all.
That is hard coding the current state of how you do things.
There’s nothing about transforming.
Okay, we save Sally or Joe a couple of hours of labor. That’s always good. That’s not we’re not debating the value of increasing the level of automation wherever you can. But let’s be sincere here, you’ve just actually made it harder to transform the process to take steps to introduce ML to have a completely different experience for your customer. You’ve just hard-coded how you currently do it as programming logic instead of Sally or Joe. That’s the point. And so, we just need to be very clear about the tactical benefits of automating tasks. And then there’s transforming outcomes. That requires a completely different approach. And it requires different tech than what is typically used to do task automation.
You warn that businesses already have too many apps. Can you explain what you mean by that? And what are the implications of that?
Yeah, this is another one where the low code, no code, guys, and by the way, we just described ourselves as a nothing like code platform. But the way that most of those platforms think is to create more apps.
If you have 50 apps on the desktop, let’s get you to 60, 65, or 70. What do you say? Well, what business leader has ever gone to IT and said, what my people need is more ugly IT screens. I just have never heard it. So we mistakenly need to increase our velocity of technology adoption by delivering apps to a desktop, so a user has to deal with it. We have equated those things, when in fact, that’s a mistake.
We need to stabilize and reduce complexity on the desktop, even while increasing the level of technology adoption. In fact, we write a really simple formula.
Your function for your adaptability technology right now is a function of the number of systems you have, the number of steps in your processes, and the number of people you have to involve in those. You need your adaptability function to be a function of how quickly you change Krista or a very robust, elegant, Intelligent Automation program. Because if you have that, your people can quickly adopt technology and transform.
I’ll give you an example this way, Kevin. We need to dissect this whole notion of how we can continue to increase the rate of technology evolution, while humans evolve over eons. But tech evolves in months, now weeks nowadays. How is that ever going to work? It works by making technology understand how people work, not the other way around. So we are conversational creatures. We’re doing that right now. Your podcast is a perfect example of this. No one went to training before they started our episode here. Right? If you go from one podcast to another, no one’s going to think that’s a testable event. Or that I need to be trained for that. We embrace change in conversation.
Krista makes all interaction with your technology. It’s a conversation between people, systems, and AI. And what that does is stabilize the desktop, which means it stabilizes the level of complexity for users. No one comes into a business hired because they’re a good salesperson, and they send them to sales training. They send them to CRM training. That’s of no value. Right? So when you join a company that’s using Krista as a part of orchestrating CRM and a dozen other systems, you don’t even know what CRM you’re using. You’re just having conversations with Krista. And the fact that you’re updating new Salesforce or HubSpot or whichever is entirely irrelevant. And in fact, it could be one today and another one tomorrow, and you didn’t even know the difference. That’s when we can get to the level of technology adoption, where IT can increase its rate of change without the desktop even knowing it happened. That is what we need, right? That’s what we’re looking for. And this is what we mean by reducing complexity.
The antithesis of that is, okay, now we have three more tools that help our CRM do things better. Give those three new tools to all our sales reps. Oh, by the way, now, we want to make inventory available from the supply chain. Great, give me an inventory tool. Now we need a new approval process for our pricing and promotions. Well, great, let’s build a pricing and configurator tool, and let’s get into, oh my gosh, we are going to die, right? Just where we’re putting robots into people then. Right? We’re not pulling the robot out of that person. We’re making people into robots. It’s like we’re serving software. Right? Now, let’s stop that.
Listeners here often tell me that it’s not the technology that keeps them from innovating. It’s actually managing the change. And in fact, the change management and the people aspects are often more complicated than even the biggest tangle of technology. What is your advice for better managing change?
They are exactly right. There’s no doubt that that is the case. Because regardless of how complex the technology itself is, that can be sorted out in a lab, stage on a server, and you’re done. The change management of humans is entirely more complicated than that. Therefore, we have to find means by which we can make technology changes that do not inflict a change management program on people.
In a previous comment I made, if your CRM is on the desktop, and the three tools you just added to the CRM are delivered to the desktop, and here comes the new pricing configurator on the desktop, you’re never going to be able to keep up with the pace of evolution of technology. Right? We all still marvel at Moore’s curve from the mid-60s and how accurate that was. And we’re seeing it even now.
An excellent example, I presented the second-day keynote of a manufacturing tech conference. Twenty new pieces of technology on day 1, 20 more right after me, right? These manufacturing firms, they’re going to want all of them. Can they run 40 Change Management Programs at the same time? Six months later, they’re all going to go to a manufacturing conference. Again, there will be 40 more tool things they want to do. We have to stabilize. And, you know what this, just to interrupt myself, our solution to this problem is to have a final assembly step from IT capabilities to the people who consume them. We currently do not do that. That’s why I keep using that analogy of the CRM plus four tools. Well, there they are sitting on a desktop.
We need a final assembly step Kevin, and what that is, we embody that in Krista and Intelligent Automation. Krista interfaces with all of the capabilities. Add as many as you want. Your change requirement is to change Krista. She’s much more comfortable with change than most people. Once we have that, people’s conversations will just naturally vary. No one ever walked up to me and said, you didn’t say it the same way the last time we talked? Well, of course, I didn’t. Because it’s a new day and things change. We embrace change. We just have to do it in a conversational context. So simply put, you want to manage change, you’ve got to separate IT capabilities. And you’ve got to have that final assembly step so that people aren’t actually impacted every time a change is made.
What are three quick action items that our listeners can do to begin to take advantage of your ideas and advice?
Here is a little mental exercise. I use the analogy of Microsoft Excel. So today, we use Microsoft Excel throughout the business to do all kinds of fun stuff. What if Excel was actually held in IT, and you had to ask IT to make a change anytime you needed a spreadsheet? We’d never get any spreadsheets, we never get any changes to spreadsheets, but that’s how we run our business now. That’s how our digital transformation works now. Well, of course, there are struggles. So I think that’s an important mental exercise. The fluidity, interactiveness, and how many ways we can change and try things in Excel are how we need all of our businesses to function and all that IT systems. So I would use the Excel analogy on myself or my peers or my boss, and I think, don’t we need to be able to change our business the way we can create and change spreadsheets? And why are we asking it to maintain all our spreadsheets in the process of automation? That’d be the first.
The second is, and this is a big one, there’s a lot of task-level thinking right now. I would ask people to think about outcomes. Outcomes are what we’re looking for. Remember, we talked already about the difference between task versus digital transformation. Digital transformation comes from a much better, more effective, higher customer set, whatever it is, lower cost outcome. So the shift from task to outcome is big. It ends up crossing departments. Tasks are simple because they are within my scope of control, right? Outcomes mean I got to deal with my peers, and other organizations mean I got to deal with more, but outcomes are a big deal.
Third, watch an episode of Star Trek. Certainly, any of them are great, but try an episode of Star Trek Next Generation, and consider how the humans in that show are talking to humans, an Android or the ship’s computer, without ever switching context to oh, wait a minute, I’m talking to an Android. So I have to think this way. Or it’s the ship’s computer, so I have to say it in a different way. Or, I’ve got a different UI to look at in order to complete another outcome. People are more effective when a conversation is how they interact. Of course, with people, but also with machines. And that’s the thinking we need to get pervasively implemented in our organizations.
What should executives and business owners be thinking about now and strategizing for today in order to be prepared for the world in five years time?
Yeah, yeah. So that is a big one, actually. We just talked about how change is something humans are not all that good at. And here comes a massive change. Today, the world is 80% doing and 20% thinking. And we already talked about how the world is moving to 80% thinking and 20% doing. Any business leader is going to have to help his organization or her organization make that transition. Everything is going to change. What we need from our people will change the growth in the skills required in organizations will change the number of people in certain roles will change dramatically.
When very little manual labor is done, and by manual labor, I mean even mean keyboard and mouse, right? Any manual labor is done, and very little of that is done by humans. That is a completely different algebra on what labor requirements are.
This gives us, of course, an enormous opportunity. There will be great winners. There will be horrible, horrific losers. I mean, an organization–people. Just like typesetters don’t exist anymore because desktop publishing killed that. And just like we’ve always had this transition. They’ve never been simple, but they’ve always ended up with a better world on the other side of it. We’re going into a massive change.
Now, intelligent automation means a whole lot less human labor. But that also means labor freed up to do much more meaningful things. So we’ve got to shepherd our organization from today to that place, and I think that’s a huge task. But I also think it will happen within the next five years, and that will be all the better for it.