Marina Gorbis's Blog, page 1444

Three Imperatives for Good Project Managers

We all have good ideas. The hard part is making them happen.

HR departments love handing out worksheets on how to plan your time. Management gurus have written hundreds of pages of advice on how to better manage your week, your day, and your hour.

But what happens when you try and apply that advice to your team or organization? How do you deliver complex, multi-level, multi-year programs of work across teams and business units that may be, quite honestly, in chaos?

I sat down with three senior executives leading major projects at complex, fast-moving organizations  and asked them what they do to be effective. They had all worked across multiple industries in their careers, run successful projects (and also some that had failed), and extracted lessons that made sense at every organizational level.

Here are the three tips they said had the biggest impact:

Be strategic, not tactical. The biggest mistake most managers make is that they spend their days and nights focusing on tactics when they really need to think about strategy. Tactics are important. They relate to the details, the fine-grain specifics to execute against a tight deadline. But strategy asks whether the milestones being created are the right ones. In the case of one media business, teams were overweight with technical project managers and lacked strategic thinkers who could prioritise what mattered. The results were hundreds of milestones that related to product launches without an overall picture of which products were the most important to customers. Strategy is all about making trade-offs. In a world where you can’t do everything, what are the one or two things you really need to focus on?

Talk about the red. In most organizations, managers spent their time sitting in steering committees hearing about how great things are going. That’s a waste of time. Talking about “red” is nearly always a more useful conversation than talking about “green”. In one government department I talked with, senior executives realized this meant changing the culture around failure. People were so afraid about losing their job that they weren’t able to do their job properly. When senior leadership started normalizing failure, employees felt more comfortable flagging when things were going off track –and working out how to fix them. Let’s be clear: it can be pretty demoralizing talking about failure all the time. Senior leaders need to know how to thank people who use their time well, and focus on the things that matter –especially the bad news.

Have leading, not lagging, indicators. Red is good. But amber is much better. Many managers aren’t able to tell their bosses if a project is off target, over budget, or past schedule until it has actually happened. Far more useful is to have lead indicators. These are triggers built into project plans so that managers have foresight into what’s going wrong as it’s happening. The key thing here is to focus on goals that really matter. If budget savings are what you care about, create lead indicators that phase in incremental savings. If being on schedule is the highest priority, create lead indicators that focus on completion to deadlines. In a cost-savings program in a major mining company, one manager built an intricate plan that culminated in big-buck savings in the final milestone. That’s no good. Much better is to break this final milestone into mini steps. Building good lead indicators means thinking about what really matters and giving that issue sufficient visibility.

We all want to be effective in our organizations. But ask yourself: are your current tools working? Following these three principles is a good step towards turning good ideas to great execution.

How Technology Creates Jobs for Less Educated Workers

Conventional wisdom holds that new technology requires highly educated workers. There is little doubt that new technologies have taken a heavy toll on less educated workers not only in manufacturing industries, but also in routine white-collar jobs. In many cases, these workers had accumulated valuable experience that has now become obsolete.

Yet it is a mistake for managers to assume that they need to hire highly educated workers to handle new technology; employees gain much critical knowledge about new technologies through experience on the job and such learning often does not require a high degree of education. Managers need to understand the role of technological maturity, the value of experience, and how employees’ technical skills develop under different business models. Indeed, economic research shows that new technology increases the need for more educated workers at first, but, as technology matures, less educated workers are hired in general.

Consider, for example, the licensed practical nurse (LPN, also known as licensed vocational nurse in some states). Many hospital managers have stopped hiring LPNs, arguing that they lack the education needed to handle new technology. The American Nurses Association has urged replacing LPNs, who require from 9 to 18 months of training depending on the state, with associate degree nurses who have two years of college training. The Association argues that technology has made the LPN position obsolete.

But a strange thing happened on the way to obsolescence. From the 1970s through the 1990s, the number of LPNs remained flat or fell slightly, depending on which statistics you look at. Since 1999, however, the number of LPNs has risen nearly 50% and wages have grown substantially. The reason: a combination of new technology and a new business model. New technologies, including advances in in electronics, fiber optics and anesthetics, allowed the widespread adoption of techniques for minimally invasive surgery. Using these techniques, surgical patients recover quickly enough to return home the same day, avoiding an expensive hospital stay.

Much of this technology has been around for a long time. For example, endoscopy has been used since the nineteenth century. But recent improvements have made the technology much better and, with that, a new business model emerged, the “ambulatory surgery center.” Centers specializing in just one type of surgery—knee surgery, eye surgery, etc.—grew much more proficient because the surgeons and other healthcare providers learned through experience. This meant that medical outcomes improved while avoiding the extra cost of a hospital stay and the complications that tend to arise from more invasive procedures. Better quality at lower cost opened up a profitable opportunity and, once insurance and government reimbursement was changed, this opportunity generated explosive growth. In 1983 there were 239 freestanding ambulatory surgery centers in the US; by 1996 there were 3,300. In 1983, 380,000 ambulatory surgical procedures were performed at freestanding centers and at hospitals. In 2006 there were 53 million.

But this new business model demanded different sorts of skills. Management experts Clayton Christensen, Jerome Grossman and Jason Hwang label the transition from the hospital to outpatient surgery as an instance of a more general transition from “intuitive medicine” to “precision medicine.”  Hospitals treat all sorts of patients with all sorts of symptoms. Many diseases, however, are difficult to diagnose with certainty, and once a diagnosis is made, not all therapies work for all patients. Medical professionals often have to make a tentative initial diagnosis, start a therapy, and then possibly modify both the therapy and the diagnosis depending on how the patient responds. This is “intuitive” medicine, and it requires highly trained professionals to make complex judgments about the patient’s condition. In this situation, LPNs can assist the medical professionals by performing routine tasks monitoring and caring for patients, but there is little the LPN can learn because the patients differ so much from one to the next.

The ambulatory surgery centers, by contrast, work in specialized areas where diagnoses are well identified, patients are screened for complications, therapies are well known and medical outcomes are predictable, if not always successful. Physicians can reliably diagnose nearsightedness or carpal tunnel syndrome, and treat these ailments with laser eye surgery or endoscopic hand surgery, respectively. The procedures are standardized and the outcomes are predictable.

This is precision medicine and LPNs play a different role. Because the procedures are standardized, an LPN learns valuable skills on the job. Because specialization limits the range of circumstances they encounter, LPNs can learn faster and better, acquiring skills that, though narrow, are more valuable. Through experience, an LPN can better anticipate the needs of the medical professionals, can monitor and care for the patients more effectively, and can identify signs of impending problems and alert the medical professionals.

In this environment, LPNs learn on the job. In the late 1970s, an LPN with 15 years’ experience earned only 11% more than an LPN with less than five years. LPNs apparently learned little that was valuable on the job then. But today, the average LPN with 15 years’ experience earns 37% more than one with less than five years. Experience has become quite valuable, even though these workers have relatively little education. And most of the growth in jobs for LPNs has been at ambulatory care centers.

This pattern is more general within the healthcare sector as the trend to precision medicine has grown, aided by new diagnostic and preventive technologies. Increasingly, doctors and dentists are performing a smaller share of the work and a variety of mid-skill providers, from LPNs and dental hygienists to nurse practitioners and physician’s assistants, are performing more. Over the last two decades this shift has created two million new jobs for mid-skill healthcare providers, beyond the growth arising from overall expansion of the healthcare sector. Moreover, the value of experience has risen for these occupations, just as it has for LPNs.

In other words, the trend in healthcare is to shift work away from the most educated workers toward less educated workers, contrary to the conventional wisdom. Mid-skill healthcare providers have important education, to be sure. Nevertheless, managers need to understand that the benefits of new technology are often realized through new business models that depend on learning on the job.

Indeed, many of the hospitals that attempted to eliminate LPNs in recent years have since backtracked, complaining of a shortage of highly educated nurses. As Peter Cappelli has found,  managers across a wide range of industries face similar skill shortages. The healthcare example suggests that rather than complaining about a “skills gap,” managers either need to step up to providing training or they need to re-examine their business models.

Do You Need to Hear a Musical Group to Judge Its Quality?

Research participants who watched silent videos of chamber-music ensembles were 26% more accurate at guessing which ones had been winners of past musical contests (such as the Saint Paul String Quartet Competition) than people who had watched both audio and video of the groups, says Chia-Jung Tsay of University College London. Participants who listened to audio without video were the worst at guessing the competition winners. In evaluating the groups without sound, the viewers were apparently responding to what they perceived as strong leadership and indications of group unity, such as the players’ proximity and similarity of appearance—probably the same factors that had influenced the judges of the competitions, Tsay says.

The Fight Over Tesla Shows How Little Value Dealerships Add

Last week New Jersey started enforcing a ban on direct sales by Tesla Motors of its path-breaking model S. Tesla’s direct sales have also run into hot water in a number of other states: Ohio lawmakers are debating a ban on Tesla’s direct sales and Texas, Arizona, and Virginia are also opposed. Proponents of a ban on direct sales claim that they are acting in the interest of customers. But is it the interests of customers they’re following or rather the bidding of the powerful car dealership lobby?

Car dealers and more generally intermediaries represent an extra layer of companies in the supply chain that clearly increases costs to customers. But in many cases they also serve important functions in a supply chain and can create more value than the inefficiencies they cause.  Let’s look at whether these functions are needed in the car market:

Search and discovery: In the same way eBay helps turn one person’s junk into another person’s collectible or AirBnb makes your empty guest room a hotel room, intermediaries can help buyers find sellers. Beyond finding the right seller, an intermediary might help match the buyer to the best product for her, providing important guidance and product information. In disaggregated markets with many sellers and buyers and where search costs are high, intermediaries provide considerable value.  But the car market is no longer like this.  With product information widely available on the Internet, numerous knowledgeable automobile blogs, websites, most customers can investigate different cars themselves. Further, the test drive can be done via a few product galleries (as is the case with Tesla). The rest of the dealership infrastructure is, basically, redundant from a search and discovery perspective.
Relationship management and trust: In markets with many small unknown sellers and buyers intermediaries can create trust between the buyer and seller and help facilitate transactions where none would have happened otherwise. The Hong-Kong based intermediary Li and Fung’s entire business model is based on orchestrating trust based relationships between garment factories in Asia and other low cost locations and established clothing brands. Again this is not the case with the auto-industry. Most sellers are very established companies with strong reputations and do not need local dealerships to help establish trust with the customer.
Local inventories: Intermediaries can hold inventory in order to provide instant delivery to customers. This was a major role of car dealerships when cars were made to a few limited standard specifications. Today, cars can be increasingly customized. In most European markets a majority of cars are made to order: a customer places the order and the car is then manufactured to exact specifications and shipped directly. This is increasingly happening in the US with higher-end cars and also with cars popular with the younger generation that prefers an individualized product rather than a cookie-cutter vehicle.
After-sales service: Historically, car dealerships and intermediaries have been an important interface for maintenance of cars. But cars have become much more reliable than they were when most of the dealership network was set up. A new well-engineered car rarely needs any special maintenance in the first 7-8 years of its life. And with a large established certified network of repair shops, a car dealership is not necessary to provide the limited maintenance a car might require.

To sum up, our assessment of the ways in which you would expect dealerships to add value suggests that there no pressing case to be made for protecting car dealerships from market forces. In the same way as when travel agents were disintermediated by websites and direct sales, the quintessential auto dealership may now be counting its last days, only creating more inefficiencies in the supply chain than the value it provides.

The process of intermediation and disintermediation in different industries is an essential part of innovating business models. It is these innovative business models that have created industries where none existed before and created immense value for customers. State governments in Texas, New Jersey and Virginia by intervening in the process are not only retarding the innovative processes at the heart of new business creation, but may also be harming their constituents.

Identify Your Primary Customer

Robert Simons, Harvard Business School professor, says companies still struggle to choose the right customer. For more, read his article.

Download this podcast

The Crisis Management Lesson from Toyota and GM: “It’s Our Problem the Moment We Hear About It”

Delay in confronting crises is deadly. Corporate leaders must have processes for learning of important safety issues. Then they must seize control immediately and lead a systematic response. Crisis management is the ultimate stress test for the CEO and other top leaders of companies. The mantra for all leaders in crisis management must be: “It is our problem the moment we hear about it. We will be judged from that instant forward for everything we do—and don’t do.”

These are key lessons for leaders in all types of businesses from the front page stories about  Toyota’s and GM’s separate, lengthy delays in responding promptly and fully to reports of deadly accidents possibly linked to product defects.

The news focus has been on regulatory investigations and enforcement relating to each company, but the ultimate question is why the company leaders didn’t forcefully address the possible defect issues when deaths started to occur.

On the recent regulatory front:

Toyota just agreed to pay $1.2 billion in a deferred prosecution agreement with DOJ and accept a safety monitor for failing to disclose to regulators—and indeed misleading them—about accelerators that became stuck on certain types of floor mats or because of certain elements in the accelerator itself.  The problem of uncontrolled speeding and deadly crashes began to appear in 2007, but it took four years, and deceptive statements, before Toyota disclosed earlier understandings about the causes and recalled millions of vehicles. The company has also settled class actions and has individual suits still pending.
GM delayed from 2005, when the issue first appeared, until recently to recall 1.6 million Chevy Cobalts with an ignition defects  that, under certain conditions, would  turn off the engine  suddenly and disable airbargs, leading to  a number of crashes and at least 12 deaths. The company is now facing the prospect of multi-front investigations from Congress, the National Highway Safety Administration and the Justice Department about why and how years passed before GM addressed the issue. New GM CEO Mary Barra has personally taken over the internal investigation, oversight of litigation and the response to government entities. She has said in a video to employees: “Something went very wrong in our processes in this instance, and terrible things happened.” Although, in contrast to Toyota, GM is just at the beginning of extended regulatory and possibly enforcement actions, the delay in dealing with the ignition issue is, like Toyota’s accelerator issue, likely to be a major problem for the company in years to come.

But, the deeper question is why these delays occurred in the first place. And it is on this question that business leaders should ponder whether they have robust systematic processes in place for personally leading or overseeing these threats to people and to the company.

The importance of thinking seriously about this set of issues is driven home by the failure of GM to address the ignition issue, even though the example of Toyota’s delay and lack of candor on the accelerator issue was a huge international story three plus years ago.

In a nutshell, CEOs and other top business leaders (with oversight of systems and processes by the board) must have a well-thought out approach to managing this type of health and safety crisis.

Preventive systems and testing should be in place to reduce the issues to an absolute minimum.


As Toyota and GM have belatedly done, the CEO should appoint a head of safety and rapid response teams to receive reports of serious harms to persons or property that may be linked to product issues.


Just as the general company ombuds system reports concerns to the top of the company about serious commercial, legal or ethical issues, the rapid response team should take any issue of potential consequence to the CEO or other high business leaders.


Most importantly, the CEO or top business leaders should then form appropriate multi-functional teams relating to: design problems and solutions; internal personnel and processes; duties to regulators; management of litigation; a communications strategy with various constituencies; and any other relevant functions.


The CEO or top business leaders must have prompt, periodic, direct reports until there is a good understanding of the interrelated issues. Then they must make decisions on an appropriate response. On these important safety issues, the CEO should also keep the board informed.


Both during formulation of the strategy and after, the CEO or top business leadership must ensure that all communications to all constituencies must be strictly accurate. It is better to say nothing—and develop accurate facts—than to issue deceptive or incomplete statements.


Once decisions are made about strategy, the CEO must oversee implementation to make sure, as appropriate, that it is meticulously carried out, changing systems both with respect to specific issues and more broadly as necessary, dealing humanely with people injured, and  communicating fully and transparently with regulators, media, and other constituents.

To take these fundamental steps is to pass the stress test. And the striking examples of Toyota and GM are an occasion for companies to review whether processes are in place to ensure that they are taken in the event of such a crisis.

Many commentators are speculating about whether the Toyota settlement with the Justice Department will be the template for a future GM resolution. To me, the more critical question for all companies is how expeditiously to handle these crises at the outset in order to avoid the unconscionable delay—and the searching regulatory problems that follow, as the Toyota and GM cases show.

It is our problem, the moment we hear about it.

What Does Professionalism Look Like?

When we talk about “professionalism,” it’s easy to fall back into the “I know it when I see it” argument.

For Emily Heaphy, an assistant professor of organizational behavior at Boston University, and her colleagues, this isn’t a cop-out. The notion of being seen as professional may be central to how we define success in the U.S. — and, consequently, how and why certain people aren’t able to attain it, depending on how well they adhere to social norms. In particular, Heaphy and the other researchers set out to study “one potential culturally bounded workplace norm — that of minimizing references to one’s life outside of work.”

They did this in two ways: First, they tested how people connect perceptions of professionalism to what a worker’s desk looked like. Second, they examined how recruiters from two different countries rated potential employees who referenced family or children.

For the former, they presented study participants with this nondescript cubicle:

Along with the description of a fictional employee:

Eric is a manager in his mid-thirties, who has been with his company for five years. He is married and has two kids. Eric’s performance evaluations are consistently strong, and he is considered very professional.

Participants were asked to then use a selection of stickers to decorate Eric’s office based on their mental image of what it might look like. Some stickers clearly referenced work (file folders, for example), others were neutral (a tissue box), and some referenced nonwork (children’s drawings or toys). On average, this is what Eric’s office ended up looking like:

When one word in Eric’s description changed, however (“he is not considered very professional”), his cubicle looked like this:

The differences are small but striking. While both professional and unprofessional Eric have office supplies and family photos in their cubes, unprofessional Eric also makes use of what appear to be holiday decorations, a poster of Elmo, and a Discman. It’s noteworthy, Heaphy told me, that the number of objects remained similar in both scenarios — in other words, you can’t simply say that an unprofessional person is messier.

Heaphy and her colleagues also asked participants to complete the same exercise with a female employee named Stephanie. Interestingly, they found no statistically significant differences in how professionalism was gauged based on gender:

But there was a significant difference in judging professionalism when they looked at how long participants had worked in the U.S.

This suggests that filling one’s office with strictly work-related items “is learned with experience living in the United States rather than a culturally universal feature of appropriate workplace behavior.”

So why is this important?

For one, it highlights how deeply rooted religious ideology still is in America. Heaphy and her co-authors trace what’s unique about the U.S. — “maintaining unemotional, polite, and impersonal workplace interactions” — to what’s referred to as the “Protestant Relational Ideology.” Basically, this is a theory originally developed by political economist Max Weber and based on “the need to put aside personal concerns to devote full attention to one’s work so as to fulfill one’s moral and spiritual calling.” It may sound out-of-date, but its effects aren’t. In one depressing example, a recent paper about unemployment found that “psychic harm from unemployment is about 40 percent worse for Protestants than for the general population.” In another, women with children reported receiving unfair treatment when they violated the norm that “workers should devote full time, uninterrupted hours to paid work.”

Related, but perhaps even more significant, is how cultural norms affect today’s more global business environment. Just as it’s no longer true that “work” means “being in an office from 9-5,” it’s no longer the case that Indian businesspeople stay in India, or that U.S. execs remain in the States. “Confusion about a tacit norm… is only enhanced by the growing globalization of the workplace that increasingly brings workers together across national borders,” write the study’s authors. This can “result in misinterpretations and misunderstandings.” In fact, one of the most important things people must do when working in new cultures “is to discover and respect the norms of their new setting, or suffer the consequences.”

But right now, those consequences might unfairly be pushing — or at least excluding — non-Americans.

And what if U.S. businesspeople work abroad?

“The question we don’t have the answer to is, ‘What are the ways people evaluate professionalism in other countries?’” Heaphy explained, referring to the dearth of scholarly literature on workplace norms worldwide. “In the U.S., with a Protestant Relational Ideology, you need to be completely devoted to work. We don’t have a similar theory as to why other countries would be different.” But given the chart above, we can guess that they likely are.

This assumption is further evidenced by the next portion of the same study. Researchers asked American and Indian participants to evaluate a job candidate based in part on how he or she would build rapport with a potential client. In one set of circumstances, the candidate would make reference to a photo of the client’s family; in the other, he or she would only discuss the location of the office or the view out the window. Participants were then asked whether they would recommend hiring the candidate.

While there was no statistical difference between Indian and American participants when it came to the latter example, the researchers found that “U.S. participants with recruiting experience negatively evaluated the candidate who engaged in nonwork role referencing, whereas Indian participants with recruiting experience did not.” In addition, “Job candidates’ success in advancing to the next stage of the hiring process was increased when they minimized references to nonwork roles in a U.S. but not Indian context.”

Much more research needs to be done in this area, according to Heaphy, including these potential paths: When, and under what conditions, might cultural ideologies of professionalism change? Do men and women feel as though they have different leeway in terms of displaying personal items? And do different types of workplace items or discussions — say, ones related to sports or family — illicit different reactions when people are trying to make sense of professionalism?

Though I wouldn’t necessarily reserve judgment on the colleague who still uses a Discman.

Malaysia Airlines Managers Could Have Had Better Data at Their Fingertips

Unprecedented disasters nearly always yield 20/20 hindsight, and lead to narrowly-conceived, “never again” style reforms. But sometimes the lessons are more broadly useful.  In the case of Malaysia Airlines Flight 370, the most impressive fact few reporters choose to explore in any depth is that the Boeing 777’s engine was capable of communicating long after the pilots went silent. And the tragedy is that all the information it was able to convey was not actually transmitted.

If the news media were more tech-oriented, they would report this as an Internet of Things (IoT) story. The term refers to the fact that billions of sensors and devices are connected to the web, automatically transmitting, receiving, and taking action on data. Now that, in fact, they far exceed the number of connected humans on the planet, we have an Internet of “things” more than people.

The data that eventually showed that Flight 370 continued for hours after voice transmissions ended came from satellite pings, most notably provided by Inmarsat. But more precise data could have come more directly from the plane’s Rolls-Royce Trent 800 jet engines – almost all Rolls-Royce engines produced in the recent past are IoT-enabled. While many major industries have yet to create “IoT” products, that’s not the case with jet turbines. Thanks to their cost, the very strong interest in running them efficiently, and the enormous safety implications of their maintenance, it has been easy to justify the added costs of  building in sensors that transmit information about the engine’s operating conditions to the ground in real time. All manufacturers now routinely include 60 or so such sensors on their models. (Evidently these engines, because they were older, had add-on sensors, not the built-in ones, but the point is the same.)

This data allows the manufacturers and their customers to do things that were simply impossible in the past. Instead of scheduled maintenance, whose intervals are typically based on imprecise measures such as mean-time-to-failure, they can now offer airlines “predictive maintenance,” which is done when each individual engine actually needs it, as indicated by real-time data from the engine such as detection of a hairline stress fracture that may not be visible. That’s not only cheaper, but can also avoid a catastrophic in-flight failure.

The IoT turbine also offers the manufacturer a marketing option: instead of selling the turbine, they can instead offer an airline a lease where the price is determined dynamically, based on the precise data about how far the engine is flown, or they can offer fixed-price maintenance agreements that work because the company can avoid unscheduled repairs.

So why didn’t engine data help in the search? Because the airline had the option, but declined, to pay about $10 per flight for real-time access to it. Had it chosen to make the investment, it might have done so in order to tweak the engines for maximum performance – but in doing so, it would also have gained the ability to keep tabs on a wayward plane.  Looked at in this light, the terrible delay in finding Flight 370 can be seen as a failure of management practice to keep up with new technology.

IoT experts I’ve consulted about this idea think it’s likely that the data from the Rolls-Royce engine was probably in an “open” (vs. proprietary to the company) format that would have allowed it to have been shared, as it was generated, with the airline, and even the flight controllers. That would have meant that, when voice contact ceased, they could have switched to the engine data. Because they would have seen that the plane was still flying, there might even have been time to take measures to avert loss of life. Instead, this data only surfaced days later: still critical, but not actionable.

My take on this lapse is that there’s a gap between the power of IoT technology, and management’s ability to conceive strategies to capitalize on it, which I refer to as “IoT Thinking.” In particular, managers have yet to wrap their heads around a new design imperative: That as we design things to be part of the Internet of Things, we must constantly ask: who else could use this data?  Many of the potential benefits of real-time IoT data can only materialize if we envision multiple users.

Here’s a simpler example to illustrate what I am saying. I recently saw a vending machine prototype enabled by IoT technology, mainly with customized marketing in mind. It can recognize a customer, offer her a special package (perhaps a price break on a soda plus chips) based on her past buying choices, and even adjust prices dynamically based on outside air temperature. (Someday you might regret saying, “I’d pay anything for a cold drink right now!”). However, the developers didn’t stop there: they asked what I think has to be a fundamental question with the IoT (and one we just haven’t asked in the past, when proprietary data was the route to fame and fortune): “ who else might have use for this real-time data?” In this case, they realized that the machine’s sensors could also detect when the inventory level in the machine was low. If that data were relayed (again, in real-time) to a distributor, it could be feasible for a delivery truck to have its route dynamically rerouted to replenish the machine. One data stream, many uses.

Similarly, manufacturers will be able to dramatically improve supply-chain integration and their distribution, while optimizing assembly line efficiency by building in sensors throughout the assembly line, and spreading that data to everyone, from assembly line workers to supply-chain partners, who can use it.

Given this new reality, aviation safety authorities should consider mandating that engine manufacturers in the future routinely provide the data stream to airlines and the flight controllers — and that the companies monitor it. (That’s quite simple: all they need to do is program alerts for the relatively rare exceptions when the data deviates from norms).

Smart airlines may take the initiative and demand the data now. My sources in the industry say that some airlines use this real-time data to tweak engine performance, a competitive advantage. (After the BP blow-out disaster, I suggested that the same approach would be applicable to drilling platforms. And it would be ideal in the event that the Obama Administration approves the Keystone XL pipeline: a built-in sensor could detect a stress fracture before any visible evidence, then automatically trigger a shutdown. Not a drop of oil would be spilled and the company would know exactly where to go, speeding repairs and restarting the pipeline quicker.)

The Internet of Things’ ability to share real-time data among many users simultaneously has many advantages to everyone who has access to it. Not least of these is the potential to save lives.

How Robots Will Work with Us Isn’t Only a Technological Question

Many questions remain unanswered about how humans and robots will interact, including in the workplace. We just know that many more of these interactions will be taking place, as robots continue to play a greater role in our lives. Missing from many of our conversations about robots is the role of human robotic interaction (HRI) – we tend to focus on what robots can do, more than how we will work with them.

This question is critical, and its answer is ultimately dependent on questions of design. How we design our robots will shape how we work with them. In thinking about how to design our robots, a service framing is beneficial: who are a robot’s customers? The answer goes beyond those directly involved in interacting with the robot to include additional stakeholders such as families and supporters, medical staff, trained professionals such as dieticians and personal trainers, restaurants and food service providers, and even policy and lawmakers. The presence of robots will change how humans behave in ways we don’t yet fully understand, and the choices we make in designing them will help determine how.

A few years ago at Carnegie Mellon, we developed a robot, the Snackbot, along with a snack delivery service, to explore these questions. A service design framing helped us to develop the robot holistically, rather than merely seeking to advance autonomous technology. Stakeholders included customers, others in the workplace, the robot developers, designers, and researchers, the robot’s assistants, and the people who obtain and load the snacks on the robots. The context of use and the norms of the workplace also needed to be considered. A product-service system was developed to track information about behavior and preferences over time. Knowledge about how the technology influenced human behavior and how human behavior influenced the technology was used to tune the robot’s design.

Our findings, published in 2012, only confirmed how greatly design decisions weigh in shaping our behavior when dealing with robots. First, we found that that subjects anthroporphized the robot, saying things like “Snackbot doesn’t have feelings, but I wouldn’t want to just take the snack and shut the door in its face.” Second, we found that the presence of the Snackbot caused both positive and negative ripple effects in the workplace, leading to new and different interactions between coworkers as colleagues observed each others’ interactions with the robot.

Finally, we experimented by personalizing Snackbot, so that it would speak to participants drawing on knowledge of their previous interactions. In many cases, this seems to have deepened participants’ interests in interacting with the robot over time. We saw signals of trust and rapport, for example when people dressed the robot up in hats and beads, and when a customer brought Snackbot a snack — a battery — after the robot died one day in front of her office. In another case, the robot’s commenting on subjects’ snacking history led to discomfort from subjects who preferred not to be reminded of the junk food they ordered the previous day. Clearly, the choice of whether or not to personalize the robot’s service affected how humans interacted with it.

As we consider the role that robots will play in our offices and in our lives, we must remember that their capabilities are not simply defined by the cutting edge of technology. They are also the result of the design choices that we make.

Fixing a Weak Safety Culture at General Motors

On Monday, GM CEO Mary Barra apologized for 12 deaths and 31 accidents linked to the delayed recall of 1.6 million small cars with a defect in the ignition switches, saying the company took too long to tell owners to bring the cars in for repairs. The switches could, if bumped or weighed down by a heavy key ring, cut off engine power and disable air bags.

The question I ask here is not how and why did this dangerous defect occur, but rather what kind of company culture allows passenger safety to be so badly compromised?

Let’s start with the facts as we can glean them.

Fact 1: Broken communication channels. Reports on exactly when (and how) GM’s top executives learned about the switches vary. According to one report: “The company has acknowledged it learned about the problem switches at least 11 years ago, yet it failed to recall the cars until last month.” According to another, “GM has said the issue was discovered as early as 2001, and in 2004, a company engineer ran into the problem during the testing phase of the soon-to-be-released Chevrolet Cobalt.”

Fox Business reported, “Barra found out about a review of the Cobalt in December, when she was still head of GM’s global product development.” In an article on the recall in The New York Times, Barra claimed she didn’t “know the serious nature of the defects until Jan. 31,” two weeks after she became CEO, “when she was informed that two safety committees had concluded that a recall was necessary.”

This unawareness at the top would be impossible in an organization with a strong safety culture. In fact, the single-most-important attribute of such a culture is proactive and timely voice related to failures, a topic I have studied and written about extensively.

This willingness to speak up is especially true for the small and seemingly inconsequential discrepancies that can, when unreported, give rise to catastrophic failures later. Any organization can detect big, expensive failures! It’s the great companies that detect the small ones that otherwise go unnoticed. And, when news related to potential failures is withheld as long as humanly possible, safety is the first victim. Although all too human, the tendency to withhold — to wait and see — is driven out of organizations with strong safety cultures.

Fact 2: A recall. GM has recalled 1.62 million vehicles globally.

Certainly, the right thing to do is to show concern and to solve the problem. But that’s hardly a choice at this point.

Fact 3: An apology. Meeting with reporters, Barra said: “I want to start by saying how sorry personally and how sorry General Motors is for what has happened. Clearly lives have been lost and families are affected, and that is very serious.” In a video to GM employees, Barra apologized again, saying, “Something went wrong with our process in this instance, and terrible things happened.” David Cole, the former head of the Center for Automotive Research in Ann Arbor, Michigan, and the son of a former GM president, said it was the first time in his memory that a GM CEO has apologized for a safety problem. In fact, it’s more dramatic than that. As The New York Times reported, “corporate chiefs routinely avoid talking about recalls [period] unless subpoenaed by Congress.”

Barra’s apology is thus a strong and powerful step in the direction of building a safety culture at GM.

Fact 4:  A new safety leader. On Tuesday, March 18, Barra named a new head of global safety, Jeff Boyer. She pledged to meet with him on a monthly basis and offered this description of how he would function in the organization: “Jeff’s appointment provides direct and ongoing access to GM leadership and the Board of Directors on critical customer safety issues… This new role elevates and integrates our safety process under a single leader so we can set a new standard for customer safety with more rigorous accountability. If there are any obstacles in his way, Jeff has the authority to clear them. If he needs any additional resources, he will get them.”

Creating a new leadership position focused on safety is a double-edged sword. On the one hand, it shows seriousness of purpose. On the other, it risks compartmentalizing safety — which needs to be everyone’s job.

Fact 5:  Product assurances. Barra further promised to repair all broken vehicles (and to double the number of product reviews). She has offered a $500 cash allowance to owners of recalled vehicles and asking dealers to offer loaner cars while their vehicles are in the shop for repairs. 

This builds goodwill among customers, but doesn’t directly build a culture of safety within the company.

Fact 6: An internal investigation. Barra has launched an investigation internally at GM, which is expected to take about seven months to finish. 

Learning from failure is one of the most important activities that occurs in great organizations, so this step is essential. But, of course, how the investigation is handled can make or break its usefulness. Is the point to find the culprits?  Or to find out what happened, and what the company can do to make sure it never happens again.

All the actions that Barra has announced are good. Yet questions remain — including how could someone as product savvy as Mary Barra have had no prior knowledge of these issues?

Clearly, GM’s safety culture is, and has been for quite some time, badly broken. A strong safety culture stems from psychological safety — the ability, at all levels, to speak up with any and all concerns, mistakes, failures, and questions related to even the most tentative issues. Simply appointing a safety chief will not create this culture unless he and the CEO model a certain kind of leadership.

Consider the case of Allan Mulally. Soon after being hired as Ford’s CEO, Mulally instituted a new system for ensuring that he learned about problems. Understanding how difficult it is for bad news to make it up the corporate hierarchy, he asked managers to color code their reports: Green for good, yellow for caution, red for problems. He was frustrated when, during the first couple of meetings, all he saw was green. It took considerable prodding before someone spoke up, tentatively offering the first yellow report. After a moment of shocked silence in the group, Mulally applauded, and the tension was broken. After that, yellow and red reports came in regularly.

The failure to speak up about safety and other problems is not unique to car companies. Silence and shooting the messenger remains the norm in far too many companies, and this won’t change unless leaders proactively invite and embrace messages of small, large, and potential failures. To do this, leaders need to override human nature by practicing two crucial behaviors that keep bad news coming early and often:

Embrace the messenger. In a strong safety culture, leaders understand the risks of unbridled toughness. A punitive response to an employee mistake will be more effective in stifling future news of problems than in preventing them. A company’s ability to detect and solve problems is absolutely crucial to its ability to learn about them.

Reward problem detection. Failures must be exposed as early as possible to enable learning in an efficient and cost-effective way. In an interview with the McKinsey Quarterly, Mulally illustrated the shift this entails: “For example, an employee decides to stop production on a vehicle for some reason. In the past at Ford, someone would have jumped all over them: ‘What are you doing? How did this happen?’ It is actually much more productive to say, ‘What can we do to help you out?’ Because if you have consistency of purpose across your entire organization and you have nurtured an environment in which people want to help each other succeed, the problem will be fixed quickly. So it is important to create a safe environment for people to have an honest dialogue, especially when things go wrong.”

These behaviors go a long way toward building the robust climate of psychological safety that is the foundation of a strong safety culture. Companies that have this are far less vulnerable to physical safety failures that can harm customers, employees, and communities.