Preparing for the third wave of innovation with intelligent automation

From a technology modernization perspective, the digital transformation story is a straightforward but challenging one^[1]—and growing in complexity with each passing day.

Fifty-three percent of organizations believe they are “living on the edge” with regard to a potential serious disruption in their business model. In response, 81% of organizations say that “digital transformation” is “important” or “very important” to their organization. Eighty-five percent see a failure to digitize, convert, and classify business inputs as a key transformation bottleneck.

And from a knowledge worker perspective, the story is equally straightforward—but growing in complexity and frustration.

• 75% of organizations see information overload and information chaos as a key problem.^[2]
• 46% of knowledge workers say it’s challenging and time-consuming to find needed information, and 80% say they have to recreate documents that already exist because they are unable to find the original.^[3]
• 81% of knowledge workers say it would be beneficial to see documents in context^[4].

Against this context, organizations believe that the volume of information and data will grow from X to 4.2X over the next 2 years.^[5] And they further believe that 62% of this rising tide will be unstructured and semi-structured information.^[6] For the average organization, this is a recipe for disaster.

The temptation is to view technology modernization and knowledge worker productivity challenges through separate lenses. The opportunity is to view these challenges as halves of the same core challenge—creating a more intelligent digital workforce—and realize that a major part of the solution lies in infusing content with intelligence.

RPA and other machine learning technologies offer the opportunity to free knowledge workers from the drudgery of endless cut/paste and error-prone manual processes and become key enablers of a more intelligent digital workforce and workplace. That’s the good news.

The bad news is that the raw material for these machine learning engines is locked up in most organizations in siloed repositories containing vast troves of undecipherable content. This problem is compounded by the fact that the increasing volumes of content entering organizations is handled by the same siloes and technology approaches that have created these troves of undecipherable content in the first place. 

Until an organization solves the problem of turning these unstructured information assets into data that is comprehensible by a machine, all of its machine learning initiatives will be suboptimal, and undercut even the most well-intentioned efforts to create a more intelligent digital workforce via RPA software robots. 

The first wave—systems of record

Systems of record were initially defined almost exclusively through the lens of structured data processing. Largely missing was a focus on unstructured information, which represents 60%-80% of the information in any organization.^[7]

We spent the last half of the 20th century building up this capability from rows of punch cards that could process census data to global information systems that capture every dimension of our commercial landscape, from financial transactions to human resources to order processing to inventory management to customer relationship management to supply chain management to product lifecycle management, and on and on. These are the great systems of record, and like the interstate highway systems of a prior generation, they have paved the way for an enormous economic expansion.^[8]

In the mid-1990s, organizations extended their systems of record capabilities from data-centric processes to core processes that involved unstructured and semi-structured content. This was first done through technologies described as document management, workflow, and scanning/imaging (in the PC/LAN era) and later as enterprise content management (ECM) business process management (BPM), and document capture as the internet transformed technology deployment models. The important characteristic of these deployments was that they focused on large-scale, mission-critical, document-intensive processes, typically at large Fortune 2000 scale organizations.

The initial content-centric systems of record implementations automated critical industry-specific processes like new drug applications (in the pharmaceutical industry), check processing (in the banking industry) and policy processing and archiving (in the insurance industry). Solutions were difficult to use and required lots of training, but it did not matter because the people who used these systems were records and document specialists. Implementations tended to be process-specific, complex, and expensive. This was followed by deployments focused on document-intensive core back-end business processes typical of any business—such as finance (especially invoice processing and accounts payable), contracts management, and human resources. 

While solution providers began to sell enterprise content management (ECM) and business process management (BPM) as enterprise layers of technology applicable to multiple processes, the reality is that actual deployment was often still driven by departments and individual processes. As a result, information silos proliferated.

The second wave—Systems of engagement

Two parallel technology innovations with more rapid deployment cycles than anything that had come previously—mobile and the cloud—combined to disrupt the marketplace of content management solution providers in the late 2000s. The introduction of Microsoft SharePoint and later the entry of file sync and share companies like Box expanded the market both horizontally—to a universe of small and medium-sized companies for whom ECM had previously been viewed as too expensive and complex—and also within organizations—to a broad set of knowledge workers well beyond the traditional ECM scope of records and process specialists. 

While initially dismissed by the established ECM players as “not real” ECM, the reality is that SharePoint and Box transformed the ECM market with a dramatic Clayton Christensen “good enough” value proposition that drove down per-seat costs by an order of magnitude. At the same time, the entry of a host of software as a service (SaaS) solutions, led by Salesforce, changed the economic equation for what used to be fairly complex enterprise processes. This set the stage for the next wave of content management solutions, the creation of systems of engagement.

The net result of all of this is an explosion of options for business executives. Process owners can now implement their own solutions.  There is a growing realization that business processes need to be “appified” and simplified. Usability and ease of deployment become a critical component in the success of any enterprise technology—including content management. Millennials, Gen Xers coming into organizations demanded that content solutions share the same characteristics they value in their personal—easy to use, easy to implement, and available anytime, anywhere, and on any device.

Of course, the demand for these new systems of engagement didn’t replace the need for Systems of Record. Successive information management eras don’t replace what came before—they are stacked on top of what came previously. Digital competence in an organization is cumulative and not every organization is starting their Digital Transformation journey from the same place.

Content management is but one corner of the overall jigsaw puzzle that constitutes systems of engagement, but it is an important one. In a world of digitally facilitated communication and collaboration, where almost all data, voice, and video are transmitted via the Internet, every interaction leaves a trace. This has mind-numbing implications for those responsible for securing, storing, and deleting such records.^[9]

The signs of unrest in the content space gathered significant intensity in the past few years^[10]:

• The average number of content systems in use continues to rise; the average number of systems has grown by nearly 30% over the past 5 years.
• While most organizations continue to increase the number of content systems they use, a rising portion of critical business content (now 54%) remains OUTSIDE those content management systems.
• We’ve moved beyond the cloud “tipping point.” Even for organizations that were initially skeptical, for over 8 in 10 organizations, cloud capabilities are now a key part of the solution.
• For 70% of organizations, the monolithic model of the ECM era has been replaced by a desire to consume content capabilities as needed—i.e., content services.
• Pricing and purchase models are changing dramatically from CAPEX to OPEX—57% of organizations prefer subscription pricing models.

The impact of all of this is that while the business need to automate core systems of record and systems of engagement has not changed, how end user organizations view the core underlying content management capabilities to drive this change has changed significantly.  The first generation of ECM solutions are being transformed and replaced by a demand for more modular and configurable content solutions—content services.

The third wave—systems of intelligent automation

There is now a new wave of innovation coming into the market place in the form of AI and machine learning technologies, often grouped under the broad label of “intelligent automation.” The initial focus of this innovation is on robotic process automation companies, or RPA, but the long-term impact of machine learning on process automation extends well beyond RPA. Additional technologies that can be grouped under the Intelligent Automation umbrella include smart workflows, machine learning and advanced analytics, natural language generation and processing, and cognitive agents.