IT, automate thyself. For years, IT has been leading the charge to automate processes up and down the business, but has been criticized for not automating itself enough. Manual scripting, firefighting, ad-hoc fixes and endless coding have long filled up the timesheets of IT teams.
That's beginning to change, a recent Capgemini survey reveals. Enterprise-scale automation, while still in its infancy, is not only the job of IT, but in most cases, starts with IT itself. At the same time, automation proponents shouldn't look to automated IT for ROI or paybacks. Rather, the payback comes as automation-at-scale reaches selected critical business processes.
The survey of 705 executives finds at-scale automation that stretches across multiple domains is a rare thing, with only about 16% if the organizations in the survey are deploying multiple use cases at scale, which the survey's authors define as "implementations that go beyond pilot and test projects and are adopted at a larger scale across business units, functions,or geographies. Most are focused on operational benefits rather than new revenue opportunities, or focused on rule-based tools with few having progressed to artificial intelligence automation."
The main avenues to at-scale automation include rule-based automation (ITPA/RPA), employed by 73%, and automation using natural language processing at 47%. About 18% are working with machine learning and deep learning.
What's taking so long? The at-scale automation the Capgemini authors -- led by Jerome Buvat and Marisa Slatter -- talk about can't simply be dropped into organizations for the sake of automation. Technology alone never delivered profitability and growth. Organizations need the leadership and guidance to recognize and re-organize around new opportunities, and technology brought in to capitalize on the innovation.
Among companies supporting or piloting at-scale automation efforts, 56%, are focused on IT optimization -- so the automation ethic, as it matures, is finally and firmly taking root in data centers. Within back-office functions such as IT, the greatest focus, the survey shows, is on application diagnostics, application releases, and server automation, with rule-based technologies dominating. Of those deploying automation in IT, 81% are deploying rule-based technologies.
The report cites Charl Vermeer, IT manager for architecture and innovation at Kadaster, who points out that "Automation technology has been a key enabler for our transformation nto an agile DevOps organization. Release cycle times went from six to nine months to minutes. The zero-maintenance systems management is fully automated. This has not only resulted in a higher availability of key production applications and improved application quality, but also in an enhanced end-user experience and satisfaction."
Outside of IT, however, it's still a minority of enterprises that have automated business process on a larger scale. Customer-facing processes are next in line for automation-at-scale. At least 37% of automation-at-scale initiatives are targeted at customer service and customer experience. Another 35% of executives say they have automated their procurement and supply-chain management on a massive scale.
The back and middle offices realize greater benefits from automation compared to the front. Among organizations implementing automation at scale, back-office functions, such as finance and accounting, drives cost savings of 13% compared to 7% in the front office.
It appears to take about a year for many at-scale automation initiatives to begin to show results, the survey finds. This may clash with business leaders' expectations of overnight miracles, but good things take time. The survey shows the following average length of payback for key projects implementing automation at scale are as follows:
- Procurement and supply chain 12 months
- Finance and accounting 11 months
- Human resources 11 months
- Information technology 10 months
- Sales and marketing 10 months
Implementing automation-at-scale, of course, is more than a technical endeavor. Buvat and his co-authors make the following recommendations:
- "Set a vision and design a roadmap for automation transformation."
- "Ensure processes have been optimized before short-listing them for automation and identify quick-wins."
- "Be agile - start with proofs-of-concept and minimum viable products."
- "Ensure AI needs are accounted for at the beginning of your automation journey and use AI more strategically."
- "Build a strong business case to secure management buy-in."
- "Design the automation operating model -leaders start centrally and subsequently federate."
- "Engage business first, but bring IT onboard early."
- "Focus on change management and cultivating digital talent."
- "Set up a dedicated automation maintenance team governed centrally."
- :Continuously innovate and consider automation as part of the broader digital transformation program."