Maureen Fleming is Vice President for IDC’s middle tier research programs. She focuses on topics covering integration, process automation software and streaming analytics.
Workflow is a very common term in business.
When I talk to teams who aren’t in a technical role, they’re comfortable using the term “workflow” to describe how they collaborate to accomplish work. Their definition doesn’t have anything to do with technology and has a lot to do with how they design and document a set of agreed-upon activities to accomplish collaborative, multi-step work.
When I talk to developers or architects, workflow is much more tightly defined and aligned with technology, almost always with BPM software. There is also a narrower view of the use cases.
When you look at how workflow is being adopted by business teams outside the canonical definition, there are interesting new types of use cases that expand the use and leverage of workflow software. This is especially true in the cloud, where the orientation and thinking is more externally focused.
In this next generation of workflow, benefits are not tied explicitly to staff, tasks are not the only unit of activity, and the emphasis is often on automation.
I’m not saying that how we use workflow traditionally is no longer relevant – it is even more highly relevant than ever. Just saying that there is an elasticity to the definition that is finding a new set of use cases.
We started seeing the externalization of workflow as teams began grassroots adoption of Dropbox, Box, Google Docs and other early file sync and share cloud services.
Because sharing was so easy, team members from different companies who worked together using phone and email began to use file sync and share as a common platform to share content as they co-developed it.
As work became more structured, these virtual teams began to design a common structure to their shared folders and document a workflow to organize work activities.
Then they began to ask for their shared workflow to be automated.
In automating, it quickly became clear that workflow isn’t about staff and membership in a directory service but more about a trusted teams of individuals affiliated by the work they need to do rather than the company for whom they work.
Under these circumstances, automating a workflow has to include some mechanism for assigning tasks that does not depend on a business’ directory service and also has to accommodate an ebb and flow of participants and active projects.
This new trend is about cooperation across enterprise boundaries, greater participant dynamism and the ability for teams to re-construct their workflow as projects change.
At about the same time teams began using file sync and share, marketing organizations began to realize that they needed a better understanding of their customers’ experiences.
Useful and easy to understand customer journey maps began to proliferate that provided an as-is view of a customer experience.
Similar to the evolution of BPM, marketing organizations saw an opportunity in re-engineering customer journeys to purposefully design and execute a great experience. Tools began to be offered by vendors to support this need. Products like Salesforce Customer Journey Builder is essentially a workflow for customer interactions.
Each interaction with a customer has an associated action. There is also automation between the actions that are used to execute activities invisible to the customer. This is nearly identical to the use of tasks to hand off work from one step to the next via tasks.
In fact, customers who have reasonably modern workflow automation tools could engineer their own set of customer interactions, creating their own customer journey. Rather than tasks, the actions are centered around the interaction.
There are many emerging use cases for workflow that do not involve delivery of a task to a worker or an interaction to a customer.
With the use robotic process automation and artificial intelligence, the design goal is to automate a workflow to prevent any manual task.
With the popularity of standard APIs that provide authorized access to services, workflow is increasingly an automation bridge between two or more systems. A customer interactions captured in another system, for example, may generate an event that triggers an automated workflow that ultimately delivers instruction to a third application.
For example, when a customer arrives at a destination in the morning, immediately offer something, delivered via SMS. In the afternoon, offer something else. The next day, do something nice for them, etc.
In this case, events are triggered from new information in the apps’ database, workflow is managing the sequence of events and options, and the output is going out through email or a third party communications service.
At no point is a human managing a task in this workflow.
Another example is the use of workflow to pass tasks between two other workflow-oriented applications.
In an IoT use case around predictive maintenance, for example, one company may own equipment that is serviced by multiple third parties. Algorithms running a predictive maintenance model trigger a trouble ticketing system to generate a new ticket that must be managed internally while also handed off to a third party for field service.
Because these are two businesses with their own processes, automating the handoff is challenging.
With the use of APIs and workflow automation acting as the broker, the trouble-ticketing system hands off to the broker workflow, which uses an API to hand off to the third party field service workflow. Compliance rules are embedded in the broker workflow.
All of the tasks are handled in the source applications managed by the workflow acting as a broker.
In this case, the workflow is fully automated and externally focused.
Organizations have the opportunity to gain better leverage and speed from existing skills and use workflow tools when they can anticipate the types of use cases coming from newer initiatives.
In some cases, the problems may have existed for years, but new urgency is being applied because of high levels of visibility on the consequences of the problem. In other cases, the shift in focus of new initiatives from internal concerns to external ones also provides significant opportunities for automation.
Availability of cloud-based workflow and standard connectivity to third party applications via APIs makes it easier to apply workflow automation to a next generation set of use cases.