Every enterprise is automating. That’s not the story.
The challenge most CTOs face is that automation — up until this point — has been a largely organic affair. Different organisational groups handle it in separate, often duplicative, ways. That has led to “islands of automation” in many organisations: redundant processes and tools duplicated with little regard for standards, governance, or metrics relevant to the business such as product revenue or time to market.
In 2020, CTOs will see automation challenges arise from this. Among them:
• A recognition that rote white-collar work must be replaced by robotic process automation (RPA);
• The growing need of services to support RPA initiatives;
• An automation paradox that will slow progress down;
• A collision of automation tools in different sectors;
• A need for agile strike teams that focus solely on automation acceleration.
Challenge 1: there is little value in having administrative work done by humans
In 2020, 3.9% of cubicle jobs will be removed from the economy. Work ranging from positing account ledgers, to filing claims, to calculating HR benefits will be replaced by RPA. However, it is not all bad news. Work that requires empathy and mental agility, such as cross-domain knowledge workers, will add 300,000 jobs to the economy. CTOs must recognise the right processes to replace with RPA and prioritise highly disruptive projects.
The future of self-service is customer-led automation — Gartner
Challenge 2: RPA projects are too difficult to scale
While many organisations are perfectly comfortable doing RPA proof of concepts, scaling to hundreds, if not thousands of automated bots is usually beyond most companies’ expertise. Because of this, the market for services to accelerate RPA projects into production will grow to $7.7 billion in 2020. Expect to see the largest players in North America and Europe, but don’t forget growing RPA markets like Japan, India and Australia.
Challenge 3: a paradox is flipping incident resolution on its head
For years, mean time to resolve (MTTR) has gone down as incident management has been automated. However, an odd paradox has sprung up in the last few years: once a sufficient number of routine incidents are addressed by automation, MTTR goes up. The answer for this is simple: once you automate all of the easy stuff, you are left with the hard stuff. In order to address this, CTOs will need to adopt site reliability engineering (SRE) principles as well as address observability.
5 challenges of intelligent automation at scale
Challenge 4: the market for automation tools is blurring
When automation products are first released, they focus on simple workflows. For example, in DevOps there is continuous integration (CI) and continuous delivery (CD) tools. As workflows blur, these tools come together under one banner. In 2020, expect to see (and pay for) solutions that address both CI and CD. These combined solutions will not only make integration easier, but also offer additional enterprise benefits such as simplified key ownership and encryption of code at rest.
Challenge 5: a new organisational unit is needed to accelerate automation projects
Up until recently, the role of automating has fallen to IT. While IT still plays a critical role, separate teams are being stood up to create standards, build reusable automation, and guide governance. They do not own the processes to be automated – that is the role of domain experts. They also are not a monolithic centre of excellence. These teams, consisting of robot architects, automation jumpstarters and botmasters, strike at projects and align deliverables to new KPIs. They will be the ones responsible for, and making final decisions on, upcoming automation projects.
Written by Chris Gardner, principal analyst, Forrester
