Clearing the fog around nascent technologies, such as artificial intelligence, the Internet of Things, digital twins/simulations, robotics, and virtual and augmented reality, has proven to be an insurmountable challenge for most organizations. This is part of the reason only 13% of data science projects make it into production.

Each of these terms has reached a degree of semantic satiation, the psychological phenomenon in which repetition causes a word or phrase to temporarily lose meaning for the listener, who then perceives the speech as repeated meaningless sounds.¹ Moreover, many conversations begin with the assumption that there is a shared understanding of the definition of a given technology, when it is more likely that that is not the case.

If five people sit down in a conference room to discuss artificial intelligence, for example, setting aside each individual's emotional sentiment regarding their idea of the technology, their understanding of its inner workings and application will almost necessarily vary.

Because uncertainty registers as pain in the brain, as discussed earlier in this book, and leads to further confusion, lack of clarity is one of the greatest threats to an organization and it is the responsibility of leaders to push through this uncertainty until clarity and mutual understanding is reached.

But how?

Asking technologists to “dumb it down” damages the dignity of industry and business team members. The same holds true in the opposite direction when industry or business leaders pontificate on the finer points of theory regarding their domains of expertise.

Imagine if you gathered the top experts (on anything) from three different countries in a room, and they all spoke different languages with no overlap. The latent potential would be obvious. This phenomenon happens on a daily basis in organizations around the world, where professionals are fluent in the language of business, technology, and industry, but not necessarily in the adjacent languages, despite the fact that they share a common oral and written language.

As illustrated in Figure 14.1, these experts are capable of finding a shared language, but it requires intentionality and, more often than not, facilitation. Fluency in more than one of these languages is often a key differentiator that leads to leadership opportunities, as many leaders can attest.

But personal fluency is not enough. It benefits decisions a leader will make and provides the ability to give meaningful feedback across verticals, but the ability to foster and facilitate discussions between these groups is far more powerful.

Esther is a business leader at a chemical organization, and when she was notified that a supplier was increasing its prices on a raw material, she looked into which products used that material and analyzed the projected profitability of each product given the increased expense. She found that several products retained solid margins and therefore required no immediate action, but one product in particular was no longer profitable with the increased expense. She analyzed various scenarios and determined that a 2% cost reduction would render the product sufficiently profitable.²

How she approaches the industry leader now that she has this information is critical. Which would you choose?

• [ ] “You need to find a way to cut costs by 2% on the Product X line or we're going to have to shut it down.”
• [ ] “Our supplier for Chemical X just raised their prices, putting Product X's profitability at risk. I ran an analysis, and we'll be able to absorb the price increase if we can find a way to drive 2% cost out of production.”
• [ ] “The economics of Product X are no longer viable. It's now the lowest performer in our product line, and only a 2% reduction in cost can save it from the chopping block.”

The answer is obvious when the options are presented side by side, but the first and third options are still taking place daily all over the world. These approaches can be motivated by a desire to create urgency, a desire to wield power (“I have the power to shut a portion of your business down”), or they could be a result of moving too quickly and not thinking through the impact of one's communication.

There are three points to note about the second communication: Esther did not overload the communication with irrelevant specifics (how much the prices were raised or how profitable Product X is); she summarized the business analysis and translated it into the industry leader's context (she didn't mention the specific type of analyses she used, which platform she used, how many data points she analyzed, or with whom she verified the findings of the analysis); and she used inclusive language, indicating a shared problem and mutual need to find a solution.

Philippe is the industry leader at the chemical organization responsible for Product X, who has received Esther's communication. He and his team take pride in the fact that Product X is the most unique chemical formulation in the organization, and he is personally connected to the formulation, as it is based on research he performed during his PhD studies. Over the past 20 years, they have built a strong team, fine‐tuning the formulation for the existing product and the systems and processes to drive cost out of production. Due to the influx of competition over the decades since product launch and its effect on pricing and profitability, Philippe has begun testing a hypothesis of a slightly different formulation and the potential for the development of a subsequent, more powerful product that would once again differentiate the organization in the market.

Esther's message posed three key challenges for Philippe (beyond his personal tie to the product). First, the chemical formulation is highly sensitive, leaving little room for experimentation due to the risk of significant waste or even disaster. Second, because the product will be rendered obsolete within a year if he can complete his research, he does not want to invest time and expertise in further fine‐tuning the process. Third, his team does not have the appetite for another optimization project.

Unfortunately, none of this context changes the problem for a business leader who has to account for specific margins within a quarter or fiscal year.

Philippe hits “reply” on Esther's email and drafts three responses. Which would you choose?

• [ ] “Product X put this organization on the map. You'll have to find a way to cut costs out of the business processes to account for this or take the loss until we can launch the next iteration of the product.”
• [ ] “There is a nuance and precision to the production process for which I feel like the right robotic capabilities could be applied to increase yield and reduce waste. When we surveyed vendors a few years ago, they hadn't gotten the cost‐to‐value equation quite to where we needed it to be, but I will revisit those discussions to see where things stand today.”
• [ ] “Supplier X has been increasing their prices arbitrarily since we started working with them. Let's start the process of exploring other suppliers to see if we can get the profitability equation back to where it should be.”

In this case, the second and third answers would both be valid. Philippe resisted the temptation to overload the communication with irrelevant specifics, such as the nuances of the production process or the chemical properties and potential reactions, and instead translated into Esther's context as a business leader and employed inclusive language.

Mia is a technology leader and peer to Esther and Philippe. Philippe adds her to the email thread with Esther and asks if she and her team can assist in facilitating discussions with robotics vendors and if they might be able to first qualify the potential vendors. He shares the documentation from previous vendor discussions and the requirements that were gathered at the time with a callout that some of the specifics have changed, but the information is directionally correct.

Mia leads a technology organization that includes technical resources such as technical architects, software developers, user experience designers, data analysts, data scientists, machine learning engineers, and database administrators. Since joining the organization less than a year ago, her focus has been on establishing a hybrid cloud strategy, migrating and modernizing applications, and removing reliance on third‐party vendors in favor of a build‐first strategy.

Philippe's team has the highest reliance on third‐party vendors in the organization, and Mia and her team have been acutely interested in updating and standardizing the technology supporting his product line. In reading Philippe's email, Mia sees an opportunity to map the current state and paint a vision for the efficiencies that could be gained from streamlining regardless of whether robotics can be built into the process. Mia drafts three responses; which would you choose?

• [ ] “We don't need robotics to achieve the efficiencies you need. I'll get you set up with my team and they'll lead your team through our design and build process.”
• [ ] “My team can qualify vendors and facilitate the discussions, but we're going to need updated requirements from your team. We've created a new process in the past year, and I want to make sure the documentation fits into our standard approach.”
• [ ] “This looks like a great opportunity to collaborate. I know when we've connected with your team in the past, they haven't been thrilled at the idea of new optimization projects. If a couple of members of my team and I could tour the facility and document some of the process from our vantage point, we'll be better equipped to qualify vendors and coordinate the robotics discussions, and we might catch some other opportunities to drive cost out of the process.”

The first message dismisses Philippe's perspective and asserts that Mia's team is going to take over. The second puts the onus on Philippe's team to produce work before they receive partnership from Mia's team, placing a higher value on documentation than on partnership. The third demonstrates empathy and understanding of the context of Philippe's team, acknowledges and answers Philippe's request, and suggests beginning with partnership and joint proximity to increase shared understanding. None of the three are overloaded with technical jargon or attempts from Mia to credential herself or her team.

It is relatively easy to imagine the exponential number of breaking points within and across organizations when reading through these examples. It would have been natural and tempting for each leader to assert authority, credential themselves, or focus on their individual goals. By focusing on common goals that benefit the organization more broadly and leveraging shared language, leaders can avert many of the pitfalls that create inefficiencies and organizational divides that increase over time.

Three Altitudes of Inputs and Outputs

All technology, industry, and business processes can be broken down into inputs and outputs, and therefore a shared language across organizations.

Artificial intelligence, for example, is a broad field, with disciplines ranging from reinforcement learning to neural networks to clustering and new approaches, applications, and breakthroughs on a weekly, if not daily, basis. Many data scientists have made considerable effort and taken large portions of meetings to explain the finer points of an algorithm to their colleagues. There are times when these details are pertinent to the technologist's business and/or industry peers, but more often than not, breaking it down to inputs and outputs to support the decision that needs to be made would save time and better serve the goals of the organization.

In the case of a machine learning algorithm, the input could be location data and the output could be a clothing recommendation given the current weather. The underlying technology could be anything from a statistical model to decision trees, or a random forest classifier (to name a few).

In the case of a business process, the input could be an organizational change that needs to be made and the output could be a plan with guidance for leaders, managers, and individual contributors along with required training and updated incentives. The underlying process could be months or years of research and planning, analyzing the behavior of team members, performing surveys of the organization and analyzing the results from psychological, neurological, and organizational science perspectives.

In the case of an industry‐specific process, the input could be sunlight shining on solar panels and the output could be electricity when a bedroom light is switched on. The underlying industry specifics include the conversion of energy, electrical wiring and casing, electrical currents, grounding, batteries, and much more.

The ability to break information down to its inputs and outputs is a key component of orchestrating experts across disciplines to achieve remarkable results. Figure 14.2 presents a structural framework for inputs and outputs that flows from the top altitude to the most granular. Organizations can benefit from starting with the top‐level input and output as a baseline for communication across domains. Colleagues can then drill down to the depth that is needed to inform the decision being made or provide necessary awareness.

Over time, organizations will find their specific level of depth required to be understood across all three pillars for a decision to be made, although the average is likely less deep than an expert from a given field would assume.

Some of you might have one or more colleagues who immediately come to mind as experts who struggle to calibrate their communication to the accessible with guidance altitude, much less the universal. These colleagues have mastered their craft, often steeped in research, application, and years of experience. Along the way, they can lose a sense of how little context others outside their field understand or need to understand in order to collaborate. More often than not, collaborations with these experts begin with the well‐meaning intention to ensure that colleagues have enough context, and are unsure as to whether their colleagues will be able to ask the right questions and delve into the deeper layers of a topic as needed.

Each of us is a subject matter expert in our own field(s), and the imperative is on the individual to calibrate our discussions to the most effective altitude for the purpose of a given discussion.

Below are two examples. Where would you plot the altitude of these discussions and how effective do you feel these meetings would be?

Example A

The regional director of an organization's real estate/facilities management team is meeting with a data scientist and a facilities manager. The regional director has asked the team to find ways to reduce emissions across the organizational campus footprint.

The facilities manager opens the discussion with a review of the ask, then an overview of heating, ventilation, and air conditioning systems. He shares the science behind chillers and boilers, condensers, cooling towers, water flow rates, and wet bulb temperature readings. He feels it is important for the rest of the team to understand the factors, criteria, constraints, and control variables.

The data scientist then gives an overview of her process and approach. She starts with a beginner's guide to understanding statistical methods, what good data looks like, and considerations that would inform the length of time it would take to complete a project.

Example B

A senior manager and her team at a utilities company is meeting with a team from the internal information technology organization. She opens the meeting with the goal: they are hoping to use technology to modernize their approach to asset management. She shares that their current approach to replacing parts across the electric grid is static, based on manuals written 30 years ago. She and her team are hoping that there might be a way to predict when parts will fail and send technicians to fix and/or replace them. They believe this will lead to getting more use out of parts that wear and tear more slowly, and fewer outages by replacing parts with increased wear and tear prior to failure. Multiple billions of dollars are spent maintaining assets across the grid, and reducing this by even a single percentage point would make a material impact.

The data science team subsequently asks clarifying questions and begins sharing the kinds of machine learning algorithms that could be used to solve this problem. They share that a first‐principles simulation of the grid could be created that would then serve as a testing ground for deep reinforcement learning to test various hypotheses in a simulated environment. They explain how deep learning and reinforcement learning have been put together to create a newer, more targeted discipline, and that the resulting neural network would likely be well‐suited to this problem.

In Example A, the goal of the discussion, as set by the regional director, was put forth in universal terms. The facilities manager, however, dove much deeper into the specifics of his field than served the purpose of the meeting. The data scientist started in the domain‐specific expertise with the overview of statistical methods, but then transitioned into accessible and ultimately universal. This is plotted in Figure 14.3.

In Example B, the senior manager gave enough context to make the subject matter accessible and relatable, as well as to assist her information technology peers in narrowing the scope of their questions and the partnership. The data science team starts and ends the discussion within the context of domain‐specific language.

In either example, the onus lies with the individual to correctly calibrate the altitude of their contributions to meet the varying levels of expertise within the room. In a meeting that consists solely of technical experts, for example, there is no need to venture out of domain‐specific language. It is up to each individual to be aware of the altitude needed for a given meeting.

Managers and leaders have a different responsibility. Many have achieved their leadership purview based on their ability to understand cross‐disciplinary language, confer with various experts, and make informed decisions. The limitation of this approach is that the leader's ability and bandwidth to understand, push for deeper insights, and ideate becomes the bottleneck through which innovation must funnel, especially given the time constraints faced by most leaders.

Facilitating and developing common language between these groups of experts will empower the organization with exponentially more valuable ideas and solutions, and is the leader's imperative in the era of Autonomous Transformation.

This is achieved by calibrating the altitude of the discussion, then simplifying through breaking the discussion down to inputs and outputs. The following is an example of this in action.

The information technology and manufacturing operations organizations at a manufacturer are meeting to determine how a new autonomous artificial intelligence capability might be leveraged for increased yield and reduced waste on the manufacturing line. A data scientist begins sharing specifics of the technology that the manufacturing operations leader feels is deeper than the team needs to understand. She interrupts the data scientist and remarks on his obvious expertise on the subject and asks if he can break it into simpler terms to make sure everyone can understand. The data scientist agrees, and the manufacturing operations leader proceeds to the whiteboard, draws the inputs and outputs framework, and shares an example from a manufacturing perspective: at the universal level, suppliers ship them raw steel (input), and they manufacture vehicle parts (output). At the accessible level and zoomed in on one of the areas where they have the most waste as an organization, the presence of variability in the process means that the machinery tends to get out of alignment at various points across the line (input). Because quality inspections are scheduled, sometimes they do not catch the problem until many flawed components have been manufactured that then need to be scrapped (output). She hands the whiteboard marker to the data scientist and asks if she can share what kind of technical inputs might be required to explore addressing this challenge, and what kind of outputs could be achieved.

If you decide to try this approach at your workplace, it will be important to avoid weaponizing the framework. Approaching the topic with vulnerability and consideration will be critical to building and maintaining trust with your peers or team members.

This work is nontrivial, and market leaders will be determined by the degree to which their leaders and managers master the ability to facilitate and develop common language between groups of experts to empower the organization with exponentially more valuable ideas and solutions.