Getting Hyper Automation Right

What is hyper automation?

Dheeraj Saxena: It’s not a platform, nor a technology. It’s a framework that suggests that everything that can be automated will be automated eventually. It blends artificial intelligence, automation, intelligent business workflows, as well as process transformation, robust access to data, and substituting human effort with software – without major modifications to the existing technology landscape.

What are the benefits?

Lower costs, improved productivity, and, if done right, happier customers and employees.

Can you give us an example?

One example is in banking, which has seen a long-term decline in foot traffic to branches. Many banks built intelligent platforms around 2015 to offer more self-service features aided by things like conversational AI. They were using chatbots and scanning technologies to capture, say, mortgage applications, instantly getting a credit rating and an approval decision to send to the customer. COVID made this need even greater. So many banks have now deployed hyper automated systems with the entire business logic and banking logic integrated into an intelligent, secure, 24/7 customer environment.

How widely is hyper automation being used?

There is automation, robotic process automation and then hyper automation. Automation simply automates data exchange among applications without much effect on workflow. Next comes robotic process automation (RPA) which automates routine tasks for which there are clear, consistent rules. These two areas make up probably three quarters of the automation we see today. But ten years out, I see hyper automation making up 60% of the pie across regions and vertical markets.

Why is the percentage so low now?

The biggest stumbling block is easy access to high quality data. Many organizations lack the advanced analytics required to predict an outcome based on past experience and the ability to automate the whole business value chain.

Which industries are furthest along in hyper automation?

Healthcare comes first, followed by banking, financial services and insurance. Next is supply chain and logistics, because the workflows are so complex. After that, utilities and government services and travel. The common theme driving adoption here is the volume of customers, or users, and the complexity of the workflows.

What are some of the key success factors in hyper automation?

Hyper automation is not about technology. Hyper automation is all about understanding your business workflow and business logic. We see organizations fail when they try a “big bang” approach to hyper automation but don’t understand the complete business process before implementing new workflows and platforms.

For example?

Let’s say a bank automates the credit card application ingestion process but fails to automate the processing of those applications. Automating only one part of the process creates a downstream backlog. If you don’t consider those downstream processes, you’re bound for failure, hence analysis of the whole business workflow is crucial.

What should such a bank have done?

Along with automating the application process, automate the screening and approval process. For example, if someone has a satisfactory level of savings, a good credit score, or whatever criteria you use, instantly approve them for a certain level of credit.

What should someone looking to implement hyper automation do in the next, say, three to six months?

Understand where you are bleeding, in terms of cost, revenue, customer experience and user experience. A consultant or group of consultants can help identify the low-hanging hyper automation fruit that will deliver the fastest return on investment. Once you’ve proven the value of an initial hyper automation project, you can go with a wider implementation to create organizational level value.

But don’t go “big bang” up front?

No. We see companies that say: “We want to be a digital company. Let's go automate this, automate that.” They invest a lot of money and see nothing coming out of it because they have not understood the process. Start small, build a use case. Convince your internal organization, which is going to be difficult, and then take bigger steps using the best practices and knowledge you gained hyper automating your smaller processes.

Which members of the C-suite should be involved in hyper automation?

The CEO, the COO, the CIO, the CTO, and the CFO at least. The CFO, of course, signs the checks. The decision to embark on hyper automation will be taken by the COO because he's looking at the business operation. But again, he doesn't know the technology. For that you need to have the CIO or the CTO, because they will be ultimately responsible for the upkeep of those hyper automation platforms, the integration layers and other infrastructure.

The value proposition will differ for each role. The CFO, for example, might say I’m spending $50,000 a year on new software licenses but can save $120,000 a year on staffing. The CEO might look at the increased employee productivity. And then there is the customer experience officer who would look at the benefits for the customer.

What’s needed in terms of technology for hyper automation to succeed?

The first thing you need to look at is integration. How can you link your applications together? That's number one because the major problem that hyper automation faces today is the ability of these automation frameworks to connect to your core systems. How well can you integrate the data from your applications to create a data lake or analytical engine or AI models? You need that data to predict the effects of actions and automate processes.

Second, you need to understand your internal and external application landscape, the entire range of systems that allow you to work with your customers and business partners. Application heat mapping will give you an idea of how your applications are being used, their level of integration, where they are hosted and how secure they are. Without that global view you’ll implement hyper automation only in pockets of the business, which is not going to give you any benefits.

Is hyper automation a threat to jobs?

Automation does remove some roles in an organization, and there is no hiding that fact. That can be mitigated by retraining those employees who are willing and able to move to new positions which involve more complex and customer-facing roles. This would be an optimum use of their skills and experience towards creating a more motivated workforce. For example, in the insurance world all the basic, level-one underwriting today is being done by artificial intelligence engines. Some former level-one underwriters have been retrained as actuaries, doing more niche and complex tasks. But there will always be some employees who are unwilling or unable to do that.