This analyst opinion provides a snapshot of intelligent workflows, including some forward-looking technologies Omdia expects to see within the process automation market.
The market outlook for intelligent workflows
In today’s automation landscape there are many solutions on the market: robotic process automation/intelligent automation (RPA/IA), business process automation (BPA), integration platform as a service (iPaaS), and low-code application platform (LCAP). Each solution has a unique value add to the automation recipe and each solution would ideally be positioned as the primary platform around which the others are centered. Across all these peripheral technologies, Omdia views RPA as the center technology. RPA in its current form can be defined as software that programmatically automates the performance of routine, rules-based repetitive tasks. Whereas IA is viewed as the AI-powered progression of RPA used to tackle more complex processes where RPA alone cannot solve.
According to Omdia, this market is forecast to grow more than 45% CAGR from 2020–27 to a total market size of $14.76bn in 2027. As a market summary, the COVID-19 pandemic has created a clear paradigm shift in the way businesses operate. Overall, a rapid change for the need of increasingly digital platforms and connectivity has subsequently bolstered capex resulting in a favorable market position for intelligent workflow providers.
Specifically for the RPA/IA market segment, vendors were able to take full advantage of this trend building upon high adopting markets such as financial services and IT through evolving business functions. A highlight example is the emergence of “customer experience” as a leading performance metric for RPA/IA end users. According to a recent study published by Omdia, RPA & Intelligent Automation Market 2022 study, customer experience is forecast as a future leading business function. Amongst the RPA landscape, customer experience is forecast to grow more than 32% CAGR from 2020–27. While the utility of intelligent workflows among end users is known, there still are various challenges and future innovation required to further propagate market growth. The main ingredient for the RPA market will be the increased injection of AI. This has already been seen by the significant rise of intelligent automation within existing RPA frameworks. As of 2023, it is expected that the market segment for AI will overtake traditional RPA by occupying approximately 55% of the combined RPA and intelligent automation landscape. The major reason for the preference of AI within RPA is the progression of automating increasingly complex processes. Commonly these processes will involve unstructured data that RPA alone cannot process. These data sources typically come by way of documents, emails, and natural speech as opposed to structured data forms.
An anecdotal view on the importance of intelligent workflows
To illustrate the value of intelligent workflows, I wanted to include a personal story of navigating processes in the absence of intelligent workflows. Like many people around the world, the COVID-19 pandemic was a highly disruptive time and, as a result, I underwent a massive professional journey of changing locations and roles in a time where in-person meetings and access to information was challenging.
Overall, as an employee of a global business, I had to navigate two location changes and two role changes all while access to mobility resources and information was highly disrupted. While moving regions in a remote position might sound straightforward, many of the mobility and HR specialists will know this is no trivial task. In almost all cases, each region within an organization will have a unique system and country-specific data structure allocated to various processes. Some of these processes might include company expensing, travel, personal leave, company share plans, and several other miscellaneous datasets, which if not automated are typically governed by regional managers or dedicated personnel.
As a result of various regional legislation and information transparency, navigating these tasks became increasingly time consuming. Commonly, each process would encompass several steps, and normally individual data would be dedicated to a regional or process-specific “silo,” with no option of compatibility. It is estimated that the aforementioned process took somewhere between 75–100 separate emails and approximately 20–30 calls to internal and external business partners.
Ultimately, the biggest takeaway from this experience was the individual challenges each task created, illustrating the great benefit of stratifying and combining data in an intelligent workflow. Considering a case of thousands of employees across a typical enterprise, processes such as day-to-day HR queries, financial, and IT tasks consume a large amount of time and resources often lengthening the time to achieve success.
IBM Watson HR system as a use case
As an example of best practices in the case of HR, IBM Watson Assistant is currently being used as the global solution for IBM’s HR automation. Since 2020, IBM have deployed AI virtual agent chatbots to manage over 200,000 IBM employees for common HR queries such as financing, PTO, promotions, mobility, and a variety of other use cases typically carried out by human HR.
In this use case, the key performance metrics addressed by the intelligent workflow are the number of interactions, time of transaction, error rate, and time saved. Overall, IBM Watson’s chatbots are being leveraged to address employee experience throughout HR business processes.
In 2021, IBM chatbots have achieved:
- 6.6 M interactions up 135 % from 2020–21
- 450K HR transactions automated
- 75% quicker transaction execution
- Decrease of 89% error rate
- Time spent “8-hour workdays” of effort for HR business partners decreased from 4 to 0.5
According to IBM, the next step in improving customer experience and overall automation efficiency is to further leverage machine learning training and operations for both supervised and unsupervised data. In this case, the chatbots will be able to access history for more complex queries, which will avoid the re-explanation of problems, which IBM coin as the “spiral of misery.”
Digital fabrics – Evolving intelligent workflows
While existing automation software markets are experiencing a surge in demand, there is still a large amount of unrealized value for wider ecosystems. In today’s market, a gap still exists between vendor messaging and enterprise adoption, with the average enterprise not yet ready or able to combine regional or vertical business and horizontal business functions.
Looking forward, a major caveat in the outlook of automation software markets is addressing the challenge of “data silos” and the demand for intelligent integration, unification, and protection across multiple workflows. Data silos, also called information silos, are pockets of information stored in different information systems or subsystems that don’t connect with one another.
A digital fabric is the technology earmarked as the path toward the goal of data democratization across a global enterprise. A current description of this technology is a top-level architecture that facilitates the end-to-end integration of various data pipelines and cloud environments fueled by metadata.
The inherent benefits of digital fabrics will be the reduction of data bottlenecks by facilitating self-service applications eliminating the need for IT support across an enterprise. Furthermore, digital fabrics can also increase data autonomy, as individual stakeholders will more easily be empowered to generate unique dashboards and workflows based upon individual business needs.
Many automation software providers are developing unique tools and integration to overcome the challenge of data silos, including the hyperscale’s. When assessing the current digital fabric offerings, many of these solutions look similar to a current data warehouse or lake indicating a market opportunity. Since this is an emerging technology, it is expected that more providers will release digital fabric technologies to complement and integrate into existing intelligent workflows. As such, it is likely a large amount of competition and future business for automation software providers will be dependent on the strategic tooling and vertical market architecture to create a unified view of data.
- Among a growing suite of automation software(s), RPA/IA is positioned as a favorite. From a vendor perspective, most modern end-end RPA solutions offer integrated management, governance, low/no-code builds, and support for scaling where most other technologies are still playing catch-up.
- Intelligent automation, the injection of AI, is becoming increasingly popular among automation software providers. This is being bolstered by emerging use cases involving unstructured datasets.
- As automation software markets continue to evolve and scale, enterprises will need to offer unification technologies. Digital fabrics are well placed as a growth strategy, offering data democratization to various stakeholders in an operational ecosystem.
Alexander Bourgeois, Senior Analyst, AI, Analytics & Data