Digital Strategy in the Age of Smart Automation
Implementations of smart automation technologies — robotic process automation, cognitive tools, intelligent agents and chatbots — often fail to meet expectations. Rather than being part of a digital transformation plan focused on end-to-end processes, initiatives are usually one-offs targeting siloed tasks with limited potential for broader impact. Learn how smart automation deployments can go wrong and the approaches for better outcomes. Making immediate improvements is even more essential.
An interview with The Hackett Group’s Nicolas Walden (UK Program Leader, Procurement & Purchase-to-Pay Advisory) and Paul Morrison (Global Lead, Smart Automation Practice).
BEATE HAUSMANN: Paul, would you like to start with general comments about smart automation and how organizations can use the technology?
PAUL MORRISON: The situation with smart automation today is one of massive opportunity and newly possible combinations of tools. But that also makes it confusing when approaching it for the first time. One major challenge people have is understanding how these tools fit into the bigger picture.
The purpose of smart automation typically falls into three general categories. Smart automation doesn't exist on its own: It is one of three fundamental automation levers that every purchase-to-pay function has: ERP and core systems; best-of-breed tools like cloud solutions, which are designed to solve specific problems in a particular function; and smart automation, which consists of cross-functional, multipurpose tools used to fill gaps. It's up to each organization to find the right combination of these tools and evolve that model over time to enhance the function.
Second, smart automation enables companies to reach the full potential of digital. It uses capabilities like advanced analytics to support process improvement, transformation and data-driven behavior. Done well, smart automation takes advantage of the full power of digital to drive automated and agile processes.
And finally, smart automation can be used to disrupt the way business is done. Machine learning and AI can imitate human judgment processes, such as anomaly detection and behavioral predictions.
MS. HAUSMANN: Which types of process areas and activities are organizations prioritizing for smart automation today?
Nic Walden: Much of the focus is on robotic process automation and APIs, which are used to fill process automation gaps. But there is also considerable focus on efficiency, productivity and analytics. It's common to start with areas that continue to pose the greatest challenges in purchasing: payments, invoicing and master data. Analytics can provide feedback about end-user behavior or encourage them to follow a process or policy. Another great opportunity is using analytics to prioritize the workloads of teams. Instead of team members coming in at the start of the day and simply addressing items sitting at the top of a queue, smart analytics can prioritize the queue according to the biggest performance needs, whether that's meeting SLA targets, making on-time payments or expediting orders.
There is also a productivity opportunity in tactical sourcing bots. Vendors are bringing these to market at the category level now, which presents an opportunity for buy desks or "three bids" sourcing.
MR. MORRISON: To build on what Nic just said: It's important to recognize that smart automation is very broad in scope. RPA, chat, cognitive, smart capture, data capture and orchestration are some of the best known. Each of these advanced technologies is meant to tackle different types of work and problems. Plus, each client I work with has a slightly different idea about where the opportunities are. Most often these are in transactional areas of activity — accounts payable and the invoicing process. This is where it's common for systems to be poorly integrated and require manual intersystem checking, uploading and administration. That transactional gap is almost always where organizations will start with smart automation.
As Nic said, focusing on the biggest value driver is important. Efficiency is a clear and measurable benefit in automating these tasks, but the bigger value is where you can inform business decisions to improve the way the function works and to reduce risk. Some of the best examples we see are the result of combining AI and analytics to detect anomalies, duplicate payments and fraudulent activity, or identifying patterns in transactional data and then either feeding that information to the experts and the process leads, or in some cases, making direct automated interventions.
One client, a UK retailer, recently shared that they're using automated anomaly detection to identify duplicate payments. They detected a million pounds in value from the first run in the pilot alone.
MS. HAUSMANN: Can you share some information about hurdles that organizations face during digital transformation and how they can address those?
MR. MORRISON: In smart automation implementations, getting everything right — IT engagement, access to systems, crafting the design, building everything properly, testing, training stakeholders, etc. — is difficult. It's no surprise that organizations struggle. If the wrong opportunity has been selected for automation, no amount of resources you throw at it is going to produce a good result.
Most organizations face a range of challenges. Addressing these is best approached through all-around knowledge and understanding — knowledge of where the opportunity is, of how to implement, of how to overcome the barriers, and so on. The question then becomes, where does that knowledge come from?
Some organizations have grown it organically over time from small technology pilots. Over 10 years or more, they've built up deep knowledge to shape these digital capabilities. Alternatively, they have acquired or hired expertise from outside the company or extensively trained internal resources. Either way, the road to achieving a successful smart automation implementation is filled with potholes and must be approached with care from the outset.
MS. HAUSMANN: How do you think new digital technologies will impact the future of work?
MR. MORRISON: To start, resource needs will change. Automation will reduce the amount of manual labor required. However, headcount reductions won’t be too dramatic as there will be increased demand for analysis and partnering, creating work in new areas. Similarly, there will continue to be more focus on influencing business value and having technology enable new forms of value creation.
MR. WALDEN: Like many other operations processes, purchase-to-pay technology has the potential to transform the way work is done. While significant progress has been made, there is still a lot of manual work and drudgery that can be removed, like rekeying data and other administrative activities. A great deal of opportunity remains to harness technology in a way that complements and augments human skills, supports change, and taps into new and more strategic sources of value.
by Beate Hausmann of The Hackett Group
Available for download with permission.