I recently talked with Simha Sadasiva, CEO and Co-Founder of Ushur. Ushur helps companies and their customers by offering robotic process automation (RPA) and other machine learning approaches to support work tasks like email triage, billing, insurance claim processing, and more. Ushur enables business teams (e.g., sales, customer service, claims processing) and IT experts to automate aspects of the work that goes into serving the organization’s customers. The business teams sit in the middle of the organization with both broad and detailed understanding of the work to be done. Using the Ushur user-friendly platform, the business teams can play a leading role, rather than the organization’s technology team having to take the lead. Having the business teams in the lead is an important step toward my bottom-up vision for work crafting with automation.
My current research targets a bottom-up approach to working with automation. From Siri on our phones, Alexa in our kitchen, and artificial intelligence (AI) in our work, we all can be thinking about how best (and whether) to augment how we do things. The alternative is that others will do this thinking for us (top-down), and I don’t think that’s what we want. Research suggests that people who craft their work (shape, mold, redefine their work) can increase their performance and employability.
This post is part of a series. I hope you use the comment block below to share related topics you would like to see and, especially, examples where people are augmenting their work from the bottom-up.
How Middle-Down Can Trigger Bottom-Up
While Ushur’s is a middle-down approach, I expect that the diversity of tasks and roles Ushur’s tools affect will serve as a training ground for the bottom-up work crafting I’m hoping for. I often mention the ubiquity of the personal computer as being key to triggering the third industrial revolution. My take: as more people had computers, more people came up with transformational ways of using computers, and more other people saw the possibilities for their personal situation, and now the phase, “there’s an app for that” is a meme based on truth (as well as an Apple trademark). The transition from costly mainframe computers to something many people could afford to experiment and play with gives us the computing-rich environment of today.
Extrapolating from one of the examples Simha described, a mom having her second child may find the process of applying for short-term disability vastly improved from when she had her first child. Texting with a chatbot running the process allows her to take on the task on her own schedule (versus traditional business hours) with faster results. The insurance company workers see their customer satisfaction scores go up at the same time that their efforts can focus on more nuanced customer needs rather than data entry. Everyone involved may start to think about other tasks they’d like to turn over to automation, the steps they saw in implementing the initial application, and even how they would improve that process. Whether your favorite idiom is “many hands make light the work,” “two heads are better than one,” or something else, the more of us thinking about and experiencing the costs and benefits of automation in our work, the more possibilities we will all have.
As more people experience how automation can improve outcomes, more people may come to envision bottom-up work redesigns of their own and thus contribute to a richer environment as we move into the fourth industrial revolution with cyber-physical systems, machine intelligence, and the like. To make this transition well, we need all the help we can get. I believe that doom and gloom job outcomes mentioned by some are more likely, the more top-down the process remains.
Training Wheels and Triggers
I see Ushur’s “drag and drop” approach to mapping customer interactions as training wheels and triggers for more people to effectively trace the process of their work. Research on job crafting shows that, without professional training in work design, people typically adjust work in narrow, repetitive ways. Templates can serve as triggers to seeing an opportunity for your particular setting. An intelligent drag-and-drop tool can offer support to developing the work process understanding we need before reinventing work with automation.
Talent, Technology, and Technique: Building Acceptance of Text-Based Interactions
Another focus of mine is how to help people “Think in 4T.” That is, how we help people understand that our outcomes are multidimensional -- we need to have a target and then work towards that target through integrations with our talent, technology, and technique. One of the first questions I asked Simha was whether he thought the growing use of texting in our personal lives was helping in the adoption of text-based interactions with customers (a shift across talent, technology, and technique). “Spot on,” was his answer. People are texting more in general, and the text-based approach can help with human and computer understanding given you don’t have to work with different accents or audio barriers. Ushur can support 60 languages in text (voice-based assistants: Siri 21 languages, Google Home 7, Alexa 7).
Having a Realistic Appreciation for Targets
I raise this last issue as we apply automation because it is supposed to help our work, but not at any cost. Simha noted that the most effective near-term approaches to automation work with an organization’s existing systems: off-the-shelf tools, homegrown, and/or legacy systems built during a more siloed approach to work. I agree that we’re better off working with what we have than waiting for wholesale change. I also recall cases during the heyday of business process engineering where wholesale change was sometimes a recipe for disaster. Working better with what you have is an excellent starting place. It may also be another way to keep the opportunities open to bottom-up approaches as less authority is needed to work with existing systems.
Summary and Thanks
I come away from my discussion with Ushur’s Simha Sadasiva with several new (at least to me) insights:
Middle-down may be a great starting point for bottom-up applications of automation. Organizations have more experience with top and middle-down adjustments to work, and the examples may help to inspire individuals to make these initial efforts even better. It’s also easier to keep a change going than to start it from scratch.
Templates can support us as we learn to map the process of our work
Ugly prototypes (in this case, using available systems even if they aren’t the most modern) have advantages
Finally, many thanks to Simha for sharing Ushur’s approach and letting me work through my thinking on bottom-up applications of automation.