Disruption; Working from Home

Laurie Sirois, CPCU, CMSQ

June 2020

Disruption; Working from Home

We’ll need all available intelligences to navigate into the future. As we have bounded into unknown territory through the shared experience of a pandemic, understanding how intelligence factors apply to the global ‘working from home’ experiment allows a model for risk avoidance, improvement and success. Classical human intelligence is enhanced by emotional intelligence, artificial intelligence and data. This combination is needed to succeed in technologically enabled, distributed work environments. Let’s consider three elements, then look at impacts and outcome.

1 - Risk and Quality Management

Like enterprise risk management, software quality management encompasses goals supported by processes that help stakeholders decide what level of risk retention they can afford vs. transferring to a third party. There is no perfect software, just as there is no perfect human, corporation or asset class. The “Cost of Quality” (COQ) is sometimes misunderstood to be the cost of preventative measures designed to ensure that technology-enabled solutions meet requirements, are “fit for purpose” and will not cause business or personal losses (all of that should be included in the costs of production). The pure definition of COQ is everything that happens post-release – including maintenance, support, rework, opportunity costs, lost sales and other impacts that are difficult to estimate, such as loss of reputation or potential liabilities. Possible behaviors and problems are infinite; and the concept of complete coverage in both insurance and in software testing is a myth (or at the very least, extremely cost-prohibitive). To make the best decisions, people use a combination of data and a very human trait: trust. The need to “influence” stakeholders has motivated my research in emotional intelligence.

Both people and perils are unpredictable. Being able to plan for future problems based on past trends should not be mistaken for precognition. In other words, no predictive model is a crystal ball, nor should every possible behavior of human customers be tested. Those possibilities are endless, and R&D would be too expensive: we would never “ship” software. But what other mitigation tactics exist? To start, we can build in quality attributes like reliability and robustness, as well as low-cost maintainability (the ability to patch & fix defects, fill security gaps and release new features easily). The caveat is that classic quality attributes have a cost/benefit that is understood to mean you can’t have it all. In some cases, for example, high performance metrics mean that a software solution will not be as feature rich. Again, it is people who make decisions, and they will make those decisions based on a combination of trust and data.

2 – Emotional & Artificial Intelligence

After decades of research in emotional intelligence, a “mixed model” has emerged: combining a trait-based approach with the ability to identify and understand emotions in yourself and others, and to be able to use this awareness to manage behaviors and relationships (Greaves & Bradberry, 2009). The concept of “emotional intelligence” has now reached such a level of prominence in corporate and cultural conversations that the World Economic Forum predicted it will be one of the top ten skills for employees in 2020, in comparison to 2015 when it didn’t make their list.

Other skills from 2015 such as “flexibility” and “quality control” are dropping off the top ten, as machines begin to make decisions for us. “Negotiation” is still holding a spot, but not only are computers capable of this skill, they are already used to train humans in it. (Ortega, 2002)  In fact, a survey done by the Forum’s Global Agenda Council on the Future of Software and Society shows that artificial intelligence will become part of corporate boards of directors by the middle of this coming decade. [i] What will it take for mutual trust to be established between AI systems and people?

3 - Artificial Emotional Intelligence

This is a much newer term and pioneering field of study. At its simplest, Artificial Emotional Intelligence (AEI) describes inanimate machines equipped with programs designed to recognize human emotions and respond appropriately. We are already being assisted by non-humans via our smart devices, in vehicles and in a myriad of both industrial and highly personal fields - even health care. Because it is context-driven, we are more accustomed to artificial intelligence than we realize, and applications of AEI involving large amounts of data and IoT infrastructures - capable of deep learning - are in production today. But can we solve potential ethical and quality problems, and can machines outperform humans in identifying how we feel? [ii]

The “Work from Home” Experiment

In 1993, the management guru Peter Drucker presented that “commuting to office work is obsolete;” however, that vision hadn’t crystallized as recently as one year ago, when half of global companies still prohibited remote work (Grant, 2020). How could quality professionals convince employers to invest in adequate performance testing at scale when policies prohibited it? The coronavirus-induced global pandemic and resulting travel bans, shelter-in-place and social distancing mandates vastly accelerated business acceptance of the requirements for working remotely. Excluding our brave “front-line” workers across many sectors (and countless others who became unemployed); knowledge-workers who were privileged enough and now required to work from home were empowered to embrace it, regardless of how they felt about it - or whether their collaborative tools were up to the task. It was supposed to be temporary, why not try?

The resulting questions were endless: would our systems scale? Would people be as productive? What would we lose from being unable to collaborate in person? Is it possible to be socially aware when you can’t be in the same physical space as other people?  Even if a global pandemic was somewhere on a corporation’s disaster recovery plans, what some call “hope creep” meant that it didn’t make sense to over-invest in the details of execution or to over-sample the testing of this experiment – until we had to. Decisions and tactics required to enable employees and management to be productive unfolded very quickly, and the timeline extended in real-time. As of this writing, many Americans have been working from home at least a full Quarter. Employers face new risks associated with traditional office spaces and social-distancing requirements, and many will allow employees to remain remote for some time to come. Return-to-office plans may include rotations, so that only a percentage of staff are “in” on a given day. If teams and individuals can remain successful and happy, there may well be a permanent change in the office occupancy needs of many corporations.

Remote Work Challenges

A special set of challenges unfolded in all sectors, private and public, to manage staff now showing up through a computer screen. It took significant IT cycles to solve employee needs for adequate home workstations that included satisfactory and secure connections to resources. It also took many HR cycles to ensure staff felt safe and empowered for success. Technical obstacles required training on collaboration software and helping people set up their own workspaces. Yet despite the technical challenges, much of the early popular advice for successful remote work was in fact human focused. Common themes included: having a designated workspace, keeping a routine, setting boundaries and scheduling time off, getting exercise and fresh air (if possible).

Cloud-based systems held many questions.  Would the performance, load and stress capabilities scale well, given they were likely tested under lower models of sustained usage spikes? Kudos to employees at software producers who faced the need to release behind-the-scenes fixes and new features on an accelerated cadence. New features were prioritized based on new demands, and consumers noticed they could now see more than four people on a screen, mute other participants or raise their hand in Microsoft Teams meetings; while Google Duo users benefited from the automated removal of disruptive background audio caused by poor connections.

Changing routines takes mental energy, which can lead to exhaustion and burnout. It’s concerning that burnout statistics were high amongst technology and finance professionals before the pandemic[iii]; but it’s not surprising that they spiked. Workers appreciated the relative privilege of being able to continue to be productive, but they also faced disappearing boundaries between work and home and an increasing volume of work hours. The real surprise in burnout stats was that they changed direction after a few weeks – in large part because of resilience and that key trait: adaptability. When people embraced a new sense of purpose, they felt decreased depersonalization. [iv]

Human Elements

Could emotional intelligence be the bridge to get us to the other side of this, after all? Or is technology going to be the winning team player? How easy is it, really, to maintain a sufficient level of human connection through a screen? What about failure to use webcams, or keeping the audio off? Advice I typically give in EI talks is: “listen first but talk more – not less.” Making sure people feel heard when they may not otherwise choose to take up airtime in a large (virtual) setting means that small discussions are becoming the norm. “Reading the room” has become a virtual activity, and it’s harder to tell if your colleague is having a stressful day – so it’s more important than ever to have “the meeting before the meeting.” Consider scheduling quick one-off virtual chats instead of sending long emails that might be interpreted out of context or get lost in people’s in-boxes.

Learning where people stand via one-on-one conversations, and “selling” your proposal by adjusting with knowledge and support gained in advance can result in increased actionable decision making within virtual meetings. Large meetings still have their place: always assume that if you were invited to one, you are expected to add value. Review (or request) the agenda in advance, and dial in prepared to contribute. If you plan to hide out while performing other tasks, consider not going at all, and contribute input in advance or ask how you can help afterward.

Even before the pandemic, it was important to control our self-expression or opinions that may not add value in more serious contexts. Because relationship-building takes time, continue to invest in people by focusing on what’s important, while making the extra attempt to stay positive. It’s critical to take an extra few moments to consider how your actions and (especially now) words can affect others. With more reliance on chat-based collaboration tools, pause before hitting “enter” on that comment – think about whether you could be misinterpreted, given the various and unknown emotional states of everyone who might be reading it. [v] 

The time invested in increased connection will be well worth it if it saves cycles of confusion. It’s long been gauche to “wait” on someone you weren’t sure even saw your request – I suggest making the request verbal (even in messenger form) first, then following up.  Remain open-minded, not checklist oriented. More than ever, knowledge workers may resent requests that sound like paper-pushing. That said, if you believe that people show up wanting to contribute, do their best work, and help one another, you will most effectively convey that by asking questions, showing empathy and building trust before launching into your position (and you may even find that your colleague’s opinion changes yours in material ways). [vi]

Most importantly, if you are feeling overwhelmed by the increasing influx of digital communications, make sure you have an outlet. We may not be able to commiserate with colleagues at the cafeteria refrigerator or over lunch, and we may be wary of adding negativity to their days – but be available for informal chats. These need not be work-focused – some of the best relationship-building and stress-relief could come from talking about plans for your staycation or the home improvement projects that suddenly made the top of your to-do list. Ask for or offer advice on something small but important to someone else. Since more than 50% of CIOs say communication is the biggest challenge facing employees working remotely[vii], take every opportunity to get better at it. If you embody the attitude of: “how can I help?” that becomes infectious, and you’ll find your coworkers more forgiving and flexible when your requests might otherwise feel like an interruption.

Remote Work Sustainability

Providers of professional collaboration tools could use AEI to keep up with customer demands, discover which features are most-used and which might have problems. Some of the biggest companies in software have formed the Partnership on AI in order to address ethical concerns with the collection and use of what many might view is highly personal data. This group was started by Amazon, Google, Facebook and Microsoft (among others); now includes smaller companies and startups in the field of AEI; and even more interestingly – the ACLU and Amnesty International. They aim to benefit society by formulating best practices around the use of AI technology in ways that are safe, fair, transparent and accountable as well as furthering social good. [viii]

Providers of virtual conference platforms could also benefit from emotion AI. I was pleased when TechWell recently chose to deliver the software testing conference STAREAST 2020 virtually, as I was scheduled to give my Emotional Intelligence talk there in May and didn’t want to miss other sessions. While preparing for the conference, speakers learned that STAREAST Virtual+’s chosen platform had occasional performance lags, and it was suggested that my rural internet links might contribute to unacceptable slowness the day of my session - so I opted to pre-record my talk. I did so to mitigate risk, even though it didn’t feel quite right, and I reserved the option to deliver live day-of if possible.

Conference speakers, business leaders and educators may feel nervous in front of live audiences, but the alternative is to present to a video camera with no feedback, where it’s impossible to know how things are going. During live sessions, some tools will tell a facilitator if people have their attention focused elsewhere (based on the app being out of focus), but wouldn’t emotion AI be a better indicator of engagement? This could be purely opt-in and remain anonymous - and could prompt a speaker if the audience is engaged or disinterested. Fast feedback is key, and new technologies can provide that to us. Reading the “virtual” room is within our grasp, and the attendee experience would be enhanced by a more aware host. Applying a ‘view filter’ by intelligence factor would quickly provide insights and help to identify the next option when unforeseen challenges arise. Risk mitigation would be more effective if based on a faster feedback cycle of real-time, emotionally quantitative data. Each level of intelligence can bring value in helping us navigate ever-evolving communication channels.

I look forward to a day when emotion AI and other technology can help to “read the virtual room,” know when to elaborate or when to pause. The truly disruptive potential of the newest of intelligences will be to enrich the experiences of everyone whose time is involved in what is now more than just a shared experiment. If shared experiences are at the core of human connection, then artificial emotional intelligence can help us have more meaningful ones, while never replacing the people on the other side of that connection.

Endnotes & References