Process intelligence has evolved tremendously over the past decade.
Our first Two for Tuesday video series explores the growing importance of process intelligence.
I talk with Scott Opitz, Chief Product & Information Officer for ABBYY (a leading process mining and IDP vendor). In this short, informative interview, I ask:
Do you feel in the last 6 to 12 months we've reached a pivotal point in the adoption of process intelligence tools?
Do you think that approach to process intelligence has also changed over the past 12 months from going to more kind of case management from single function departmental type approaches?
Key Takeaways:
Automation projects drove need for better process understanding
Process intelligence helps companies analyze unstructured processes
Adoption is increasing rapidly as companies digitize operations
Process intelligence is pivotal for avoiding "fast mistakes" with automation
Ad hoc customer-centric processes offer great improvement potential
You can watch the interview here or scroll down to read a (lightly) edited transcript of our conversation.
Process intelligence has evolved tremendously over the past 10 years. Automation projects often fail because companies don’t understand processes well enough. New process intelligence tools are helping companies analyze complex and unstructured processes to deliver true transformation.
Adoption is increasing as companies realize process intelligence is critical for digitization. The discussion covers factors driving adoption, including wanting to avoid "fast mistakes" with automation and needing the facts to formulate digital plans. Dramatic growth is coming for process intelligence as more companies adopt process intelligence for competitive advantage. There is debate around starting with structured back-office processes versus unstructured customer-facing processes. One area of potential is to use process intelligence to improve ad hoc processes involving customers, partners, and employees.
Marty: I know you love process intelligence and you've been following the market; developing products in that space for quite a long time. Do you feel in the last six to 12 months we've reached a pivotal point in the adoption of process intelligence tools?
Scott: Absolutely. And I think there's a number of factors that are driving this. One big one is all the fervor around automation for the last 10 years has certainly heightened awareness. I mean, companies realize they didn't understand their processes sufficiently or at least in sufficient enough detail to be able to go off and actually do the automation.
And so a lot of those projects at least are the early ones. They would deliver them and then find out that they didn't really deliver on the full promise because they miss variations and how the processes were executed. You know, some of them had unintended consequences. You know the joke was always, you make things faster, which if you're not careful means you just make your mistakes a lot faster in an automated way.
And so I think that was that was a big driver and probably the most significant of the last , 7-8 years probably more. I also see another undercurrent that's outside of that automation technology or those automation initiatives.
And I can see a parallel here whereas I think you and I are probably experienced and old enough to remember, right? We saw the mass adoption 20 years ago of BI tools that were really driven by business users in that time period because they were unhappy with the complexity and slowness of getting answers to their questions.
And so they push forward with what was then called the New Age data discovery tools, the clicks and the tableaus and spot fires and these guys and what we see here is I think a parallel where the aggressive push to digitally transform business processes in this last decade and beyond has also exposed those same types of business leaders to the fact that they don't have the facts necessary to really understand the baseline operations of how their companies work?
And so if you don't understand it, how do you formulate a plan with insufficient data and even worse, maybe some flawed assumptions. So now that process mining and more generally as we prefer to think of it as process intelligence has become mainstream, right? It's been endorsed now by leading analyst firms and companies are recognizing the essential role it can play in their overall strategy.
It's causing not only those companies to broaden their adoption, but frankly it's causing some other companies to realize that they need to catch up quickly. So I think the next two to three years will show dramatic growth in this space.
Pavlik: Scott, you mentioned delivering on the full promise initially. Do you think that the approach to process intelligence has also changed in the last 12 months; going for more kind of case management from single-function, departmental-type approaches?
Opitz: Yes. And it's interesting. It depends on where customers are in their journey, right? So we have some companies who have not really applied process mining process intelligence at all yet and so there it's kind of a question, right, should they focus more on those case management type processes, or should they focus more on the more sort of rigid consistently executed like we might think about around and accounts payable process or typical sort of back-office, finance-type processes.
So I don't think for those customers, I don't think there's a clear yes or no answer as to which one they'll start with first or which one they'll be most excited. It's probably gonna be the squeaky wheel and their operation of where they see the best opportunity.
Now historically though, to the point of your question, I think organizations really didn't have a choice. The early technology is what I think of as like the first-generation process mining tools. You know, they did tend to focus very much on the kind of rigid, more formally defined processes, right where an exception was truly an exception, but otherwise they were fairly consistent.
And they avoided more complex process types. These ad hoc case management type process types simply because their tools couldn't handle them right. So with the rise of greater capabilities for analyzing unstructured ad hoc processes, those limits are removed, right?
And so if I think about that, if I say OK, in a company today maybe where do they think they're gonna find the most benefit, I would be more biased towards what I see with them focusing on those more ad hoc type processes and if yeah, if you think about it, the reasoning is the majority of the processes executed by business are on structured more generally, right, they tend to be more freeform and it really doesn't matter what the industry is.
Marty: Absolutely.
Scott: I mean, if you're doing new account onboarding and organizations that now have strong GNRC-type requirements – your KYC; identification of your customer risk and compliance checks, financial analysis for creditworthiness; all of these things, right? They're all complex processes, and each customer might look a little different.
Even in something like insurance claims the steps can vary greatly. That’s something that you tend to think of as it should be pretty straightforward, right? An automobile accident. How complicated can that be, right?
Well, there are some fundamental questions. Was the claimant at fault? Or were they the party hurt by this situation?? How many other parties were involved? Was the other party insured (because that kicks off a whole separate sort of process).
If another party was at fault, then what are the extra . . . what’s the subrogation process where the insurance company tries to get reimbursed by the other insurance company because their party was at fault?
Or any number of other variations of the type of damage that's being done. Is there gonna be a field adjuster who's gonna check on it? Or does the customer – a lot of companies today offer the ability to submit pictures and video of the damages so that they can. In fact, there are interesting uses of AI in some cases to assess the damage right?
Or do they take it to an approved auto Body Shop? And they submit the paperwork or a quote that can be trusted. And finally if the car can't be fixed. If it's totaled, then that kicks a whole other process off around salvage, right?
So the insurance company can at least recover that salvage level, something for that. The point of this is, hey, any interaction you're gonna have with a customer or trading partner or employees, they're very ad hoc in nature. And frankly, I believe offers some of the best opportunities for improvement.
Marty: Yeah, absolutely. Customers satisfaction is across the whole journey, not just a departmental journey.
Opitz: Absolutely.
Pavlik: I’d like to thank you for joining us today on Two for Tuesday. And I’d also like to thank the audience for coming in and listening to us today. If you’d like more information about process intelligence, please feel free to go to doculabs.com or abbyy.com.