How To Supercharge Your Automation for Bigger Cost Savings and More Efficiency

Intelligent automation combines robotic process automation (RPA), artificial intelligence (AI), analytics, data, and more to create an end–to-end process that can learn and adapt on its own. Organisations that have already moved beyond piloting intelligent automation achieved an average cost reduction of 32%, according to Deloitte’s most recent automation survey. 

Here we’ll cover how you can bring next-level efficiency to your business.

Two-thirds of companies have implemented some task-based automation, which is the fairly simple automation of a single process within one business unit. For example, creating a service ticket, automatic credit approval, or automated email promotions.

According to Matt McLarty, global field CTO and VP of the digital transformation office at MuleSoft, the task-based automations used at many companies today aren’t reusable, scalable, or sustainable enough to deliver the value that’s so important right now. That’s because they lack the context and sophistication that businesses need to truly transform their business and serve customers better. 

Some examples: 

• a chatbot that doesn’t recognise a customer or their order history, and offers automated responses that don’t address the customer’s problem 
• a meal kit service that can’t say why your box hasn’t been delivered because it has no visibility into the logistics provider
• an automated HR system that’s siloed from an employee success system.

“A lot of the RPA solutions are very much focused on recording how a process works, and automating it,” McLarty said. “But if there’s an exception in that process, as there often is, all you’ve done is delay solving the problem by putting a bot between me and the person who will actually resolve my issue.”

Intelligent automation provides context around your data so you get a deeper understanding of what’s happening. The context, which gleans meaning from images, text, and speech, detects patterns and can make recommendations, predictions, and decisions. 

Real-time connected data is a key to contextual automation, whether it’s internally automated business processes, or the ways in which companies automate interactions with partners and customers. 

We are in the relative early stages of this advanced automation, but it’s on the radar of many organisations. Deloitte’s survey found that AI is the next most desirable automation technology, with 46% planning to implement it in the next three years.

Once a company has delivered a handful of automations, the dividends tend to level off. That’s because the company is not able to easily identify new processes to automate, said Joe Surprenant, sales leader across Deloitte’s AI and Data Ops practices. Companies need to mine their data and processes to uncover these new opportunities. 

“Once you’ve exhausted heat maps and the idea box from process owners, you need a hybrid approach that includes both digital discovery tools and process expertise to identify the gaps,” he said. 

The technology that makes that possible is called process intelligence. It’s the data that’s collected to analyse the individual steps within a process or workflow, and can help an organisation identify bottlenecks and improve efficiency. While only about one in five companies surveyed by Deloitte use process intelligence (PI) today — several companies develop PI technologies — it is a fast-growing sector. 

“Process intelligence has opened up the eyes of many of our clients to show them things they did not know,” Surprenant said. For example, he said, one large company used process mining technology to analyse its end-to-end direct material purchasing process. The analysis identified steps in the purchase order maintenance process which had a higher manual activity and rework rate. This insight led to the development of a maintenance automation solution using a combination of RPA, business process management (BPM) workflow automation, and  analytics, ultimately driving $40 million in annual savings.

Another retail client used PI to uncover the root cause of a disconnect between the sales, payment, delivery, and return functions within its supply chain. These insights were leveraged to create a consolidated customer journey app to quickly trace and resolve issues across these functions. This led to a 23% reduction in order returns, $46 million sales risk mitigated, and 7% improvement in net promoter score. 

Process intelligence is beneficial in three ways: 

• It digitises and accelerates the discovery of automation use cases. In effect, it automates automation. 
• It provides an additional data point for what some business users suspect may be an area ripe for automation.
• It saves companies the time and money of hiring consultants to interview process owners to identify automation opportunities.

Implementing more sophisticated, contextual automation is about much more than technology. Success requires a fundamental rethinking and re-engineering of your processes, all centred around your customers’ needs. 

McLarty, who works closely with customers on their digital transformations, suggested these first steps:

Map the entire customer experience, from end to end, how all those pieces are interconnected, and what customers need at each step. Most companies, he said, consider just one piece of the puzzle — commerce, service, or marketing, for instance — and automate just that one element. 

Real-time, connected data is a game changer in contextual automation. Next-generation customer data platforms integrate data from every customer interaction, from any system, channel or data stream, into a unified customer profile. Having this 360-degree view helps you see your customers in totality. For example, it helps you see how a service interaction with a customer impacts a marketing promotion for that customer. 

Is your technology repeatable and scalable? You should seek tools that can uncover different processes to automate in your organisation. These tools extract the metadata from the process, and turn it into an automated workflow. 

MuleSoft is the enabling technology connecting data from any system or channel. 

“The automated solution has to have connections into the predictive analytics that are synonymous with artificial intelligence and all the data sources, so it can tell you everything you need to know,” said McLarty. 

It’s hard to overstate how transformative AI-based contextual automation will be. Surprenant said it will be equivalent to the cloud. 

“We’re not there yet,” he said. “But when we get over the curve to where companies are automation-first when they design business processes, and they’re truly embracing a digital worker mindset, it will be as big if not bigger than cloud.”