A wave of intelligent process automation solutions designed to eliminate routine manual operations like data entry, document handling and workflow routing has changed the way companies will think about process efficiency and the roles of their process owners. Within the next decade, knowledge workers involved in paying vendors, filling customer orders or a range of other high-volume transactional processes are no longer going to be focused on manually handling paper and digital documents or rekeying information. Their going to spend more time and energy focused on tangible business goals like negotiating better supplier contracts, cutting costs, boosting customer satisfaction or reducing days sales outstanding.
If you’re responsible for managing accounts payable, accounts receivable or a finance department’s operations, chances are that the notion of process automation is nothing new. So what’s changed that makes machine learning and intelligent process automation so exciting?
Whether you adopt an old school, rules-based workflow system, or an intelligent process automation solution to automate vendor invoice or customer order processing, the end results are similar. Paper documents are digitized. All digitized documents (scanned paper or native digital docs) are transformed into structured data that is validated, then routed for approval and coding to the right stakeholders. Approved transactions are automatically entered into an ERP (and/or ECM system). Vendors are paid. Customer orders are fulfilled. And every step of the way is monitored and managed so that all stakeholders have visibility to the process and timely access to relevant business data for optimizing cash flow and supporting other relevant business KPIs.
Automating these processes allows companies to grow and scale their operations quickly and relatively effortlessly. Process owners no longer struggle to keep up with the pace. Instead, they are able to take on more meaningful work to plan, forecast, budget and identify new strategies for lowering costs and boosting efficiency. The go from being slaves to the process to masters of achieving the desired results.
That’s the vision, at least. But in the world before the application of machine learning and a level of intelligence to the process, AP and AR teams were still burdened with handling exceptions that couldn’t be handled by rigid, rules-driven systems. Effort that had once been spend manually filing documents and keying in data got replaced by manually creating definitions that would allow old school systems to extract the right data from documents. In many cases, process owners would manually have to handle the same sorts or errors and exceptions over and over again.
This often meant dependence in IT or development to step in and help re-work workflows, redefine definitions, or update integrations to ERP or ECM systems. Which added to the cost and effort required.
While this approach worked for high-volume processes where there was a sense of urgency, and where the ability to scale operations was a concern, this approach (even with its associated overhead costs) made sense and delivered an ROI. For processes that might require lots of manual effort, but without the volume or velocity, the manual approach persisted due to a less than compelling business case.
So what’s changed? Vendors have injected intelligent into their solutions that address the cost/effort overhead by combining the low code automation approach of RPA with the flexibility of machine learning to reduced repeated manual work by process owners who can teach the systems once and let them adjust their algorithms on the fly. Which simultaneous relieve the burden from IT to reconfigure or recode anything. In essence, you could say that intelligent systems reprogram themselves, with guidance from the process owner.
Sounds promising, doesn’t it? Which begs the question, “What do I do next?” And “How can I get started to implement a solution and deliver better results at a lower cost for my organization?”
If you haven’t gotten started yet, it’s not too late. Start experimenting now and look for low-risk, pilot projects that can deliverable measurable wins.
Until intelligent process automation became practical and affordable, business accepted that manual data entry and document handling as a necessary evil and a cost of doing business. Inefficiency was considered acceptable for complex processes where the cost and overhead associated with automation was too great to justify automation.
Today, IPA and machine learning-supported solutions can make complex processes more simple and tackle manual steps that address BOTH general business scenarios and all of the exceptions that “what ifs” that were the real barriers to automation in the past.
That changes everything.