Back in January of 2020, few people imagined how the ongoing COVID-19 pandemic would reshape both business and society. We largely assumed that life would return to normal at some point in the not too distant future. Now, more than two years later, it is clear that isn’t going to happen. Instead we’ve entered a “new normal,” where rapid and relentless change is the status quo.
To help orient ourselves, now more than ever, we often turn to the experts that are best equipped to analyze the present, as well as make predictions about the future. The Shared Services and Outsourcing Network (SSON) is a community of such experts that focuses on shared services, global business services, and outsourcing. In the most recent edition of their annual state of the industry report, a number of compelling statistics emerged that paint an interesting picture of the future of artificial intelligence (AI) and machine learning (ML).
This blog article offers a summarization of highlights from the 2022 GBS & Shared Services State of the Industry Survey, as well as analysis of the trends behind the data.
According to SSON’s survey results, artificial intelligence is 2022’s second largest intelligent automation investment priority for companies after robotic process automation (RPA). RPA barely edged out AI, with 40% of respondents citing it as their number one priority vs. 36% for artificial intelligence—representing a mere 4% lead.
Ultimately, this is an unsurprising trend. The benefits of legacy automation solutions such as robotic process automation (RPA) are beginning to plateau. This is largely because they rely on structured data, or data that follows a pre-defined model or schema, which has quickly become usurped by unstructured data, or data that doesn’t follow a clear method of organization—making it far more difficult to process and analyze. It is estimated that 90% of all data was created in the last few years, and 80% of it is unstructured.
The 2022 GBS & Shared Services State of the Industry Survey found that data digitization is the top priority for companies in 2022. Despite analysts estimating that the COVID-19 pandemic has accelerated the adoption of digital technologies by several years, there is still a long way to go on the digital transformation journey for many companies.
By digitizing existing data, companies will create even more unstructured data, and will need to find new solutions for powering automations with it, as well as extracting actionable insights from it. The second most commonly listed priority is leveraging automation platforms (37%), which will be essential as companies digitize increasingly larger amounts of data. Additionally, rethinking operating models (36%) will be essential as companies begin taking advantage of these new technologies.
With data digitization the leading objective for companies in 2022, it is important to understand how companies are choosing to approach this problem. According to SSON, AI-based machine (26%) learning is the most common initiative for companies digitizing their data this year. Intelligent document processing (IDP, 25%) is the second most popular approach to digitization, which also leverages AI technologies such as natural language processing (NLP), computer vision, and more.
Interestingly, 19% of respondents reported having not yet attempted to leverage any of these approaches to data digitization, suggesting a decent portion of businesses haven’t yet begun incorporating AI into their digital transformation, they have no need to do so, or they’re simply unaware of the benefits of AI when it comes to unstructured data processing and analysis.
The rapid pace at which AI and ML is being adopted is creating a massive talent gap. Rising demand brings with it a need for technical workers that can build, deploy, and maintain AI and automation systems. In Q1 of 2021 alone, there were 37,000 AI job postings, representing a 45% increase since Q4 2020. With 39% of companies citing a lack of technical expertise as a top reason for sluggish AI adoption, it’s unsurprising that automation skills are highly sought after.
The primary driver behind this rising demand is the massive value proposition of automation. From application development to infrastructure deployment to business processes, automation is capable of powering anything and everything. Additionally, the ability to do more work with fewer resources is attractive to many companies. Although there is no single skill that makes someone an automation expert, there are a few core abilities that are necessary for successful automation.
According to Redhat, these skills include scripting, collaboration, source-code management, Kubernetes, security, testing, observability, monitoring, and network awareness (among others). After automation, the SSON’s survey found the next two highest priority skills are process design (55%) and data management/analytics (46%), both of which can play a supporting or parallel role in automation. Designing scalable processes for standardization and optimization is difficult yet necessary for successful automation initiatives, and data management and analysis are essential for scaling AI and machine learning efforts and extracting meaningful insights and value from them.
Fortunately, low-code and no-code solutions like super.AI’s Unstructured Data Processing (UDP) platform make it possible for non-technical business users to build, train, test, and deploy AI powered automations. Our website includes additional information on how this works, and our team of unstructured data experts is available to meet with anyone interested in learning how AI could benefit their specific scenario.
With 33% of companies indicating their automation efforts are in a mature state, it is evident that there remains a lot of room for improvement. Companies still in the proof of concept (POC) stage (35%) of their automation efforts were the largest group of respondents, and an additional 9% said they are leveraging no automation at all. This survey data is a good reminder that automation technology is still very much developing, particularly more sophisticated forms of IA that involve artificial intelligence and machine learning.
To better understand the scope of automation efforts, we can turn to the number of bots companies have implemented as well as the number of processes they’ve automated. The vast majority of respondents fall on the lower end of the spectrum, with 57% reportedly implementing less than 5 bots, and 38% automating less than five processes. At the other end of the spectrum, just 5% of companies have implemented over 100 bots, and 8% automated more than 100 processes.
Although it is often assumed that automation’s primary goal is to remove humans from the workforce, SSON’s survey results suggest reducing headcount isn’t the sole (or even primary) outcome of automation. When asked how companies are using employees’ time that has been freed up by automation, the top response at 27% was “redeploying full-time employees (FTE) into the business or functions / into business partnering roles.”
Saving costs by cutting FTEs (15%) was tied with “staying with same headcount (i.e., no need to add new staff) as we take on more work/expand scope.” Although fears around ever more powerful automation technologies replacing human workers are grounded in reality to a certain extent, the extent to which this will be the primary outcome of automating more processes may be blown out of proportion. Respondents also reported a number of benefits to having more time for human workers, such as delivering more analytics services (13%), delivering more innovation (10%), doing more problem solving for business customers (7%) and more.
Hopefully this article piqued your curiosity about our AI-powered future. At super.AI our mission to automate boring work so that people can focus on what matters. Additionally, we strive to make AI accessible to everyone. We carry that spirit with us in everything we do, from the technology we develop to the resources we curate, our goal is to make AI approachable. For more information about intelligent automation and artificial intelligence, check out the following resources: