Airbus

Boeing employees walk in front of a new 787 aircraft

Talent Retention and People Analytics at Boeing

I recently read an opinion piece called “Greatest Long-Term Threat To Boeing Is The Loss Of Talent,” written by Richard Aboulafia for the industry-leading magazine Aviation Week and Space Technology. Last month, aircraft manufacturer Boeing (NYSE: BA) moved many of its defense services and support functions out of Seattle, citing competitive and cost concerns. While the author agrees with the company’s line of reasoning, he says that “it’s also important to remember that when a company takes aggressive action to move jobs and reduce labor costs, it always creates risk. In particular, key skills and experienced workers can be lost, threatening execution and company capabilities.”

This article had me thinking about what it takes for a large company like Boeing to retain talent in an increasingly competitive business environment. As the author suggests, companies in a growth phase need to focus on attracting and retaining talent, but a company in a “retrenching” phase needs to focus on costs. The analogy given is Tesla versus General Motors, respectively. This in turn reminded me of a talk I attended a few weeks ago by Brian Welle, Director of People Analytics at Google. In his day-to-day role, Welle “conducts research and designs programs that strengthen Google’s Human Resources initiatives.” One of his primary areas of research is on the “unconscious bias,” a set of factors caused by our environments and experiences that influences our decision-making capabilities. Although Welle’s primary focus is to help Google employees become aware of and reduce their personal unconscious biases, during his talk he frequently mentioned the company’s overall drive to recruit and retain talent. The underlying assumption here is that Google remains in a growth phase – but I wonder what will happen when (or even if) Google reaches a point when it needs to shift focus to cost-based “retrenching” like that referred to in Aboulafia’s article. Obviously, this would require viewing Google as a mature corporation – hardly the case given the growth in the technology industry and Google’s new monetization initiatives.

Furthermore, I wonder why a company like Boeing doesn’t have a similar human resources structure to Google. This may seem like an outlandish idea, but I feel that many of the human resources functions at Google can be replicated in the wildly different industry that Boeing operates in. Welle’s People Analytics team focuses on organizational behavior (OB) issues as they pertain specifically to Google – so why doesn’t Boeing focus on the OB issues that affect the aerospace & defense industry? In my earlier posts “The Failure of Crew Resource Management (Part I, Part II, and Part III),” I focused on the failure of an OB system in the aviation sector (Crew Resource Management or CRM). I think it would be interesting to see Boeing expand its human resources functions to address industry-wide OB concerns like CRM. In my mind, Google is able to recruit and retain the best talent because its human resources professionals are focused on remedying OB issues that affect the broader industry, such as the lack of women in technology and the unconscious bias in most recruiting decisions.

At present, the cost issue remains crucial for Boeing’s short-run competitive strategy. But perhaps a shift towards the Google human resources model could help Boeing with its recruiting and retention issues in the long run. It’s definitely something I want to look into more. In Aboulafia’s words: “Boeing management needs to remember the greatest long-term threat to [Boeing Commercial Airplanes] isn’t the cost of labor; it’s the loss of talent and the erosion of core capabilities.”

Panoramic cabin view onboard an ANA 787-800 Dreamliner

Optimal Boarding Method for Airline Passengers (Jason H. Steffen)

Speaking of “Old Dog, New Tricks,” I’ve been thinking about other problems that airlines could very well tackle without physically changing anything about their aircraft or equipment. One of these is the classic boarding nightmare that we’ve all experienced – waiting for what seems like a lifetime just to get to your seat, because everyone needs to lift their hand luggage into the overhead bins. I read an article in the Washington Post by Jason H. Steffen, a professor of astrophysics at Northwestern University, who has spent some time studying this problem. Steffen argues that lengthy boarding queues are driven by two factors.

First, the practice by most commercial airlines of charging for checked baggage leads most passengers to maximize (and often times, exceed) their hand baggage allowance. I’m certainly guilty of this – on my most recent trip, I brought a rolling hand-luggage and a duffel bag that didn’t quite fit underneath the seat as per airline guidelines, so I had to spend the time storing both in the overhead space (after all, I need that space under the seat for my legs). Anyways, it’s simple: the more stuff people bring onboard, the more time it takes to store all that stuff and get everyone in their seats.

The second factor driving the problem is the boarding process itself. Most airlines board their passengers in the following way: first class, business class, membership club members (by rank order), and finally economy class. Within the standard boarding procedures for economy class, passengers will board from the back of the plane towards the front. As a result, the majority of passengers get stuck in the aisles, waiting for those who boarded ahead of them to store their luggage and sit down. This causes the painful boarding queues that frequently extend out of the aircraft and onto the jetway. According to Steffen, “the problem is that boarding from the back to the front is a serial process: only one action at a time is completed…The aisle in the airplane isn’t used effectively.” The only other boarding process currently in service is the “industry gold standard of open seating,” pioneered by Southwest Airlines and popularized by other low-cost carriers such as Ryanair. In this model, passengers don’t have assigned seats at all, and boarding time is significantly improved.

In Steffen’s view, “a more efficient way to board would have only as many passengers in the airplane as can put their luggage away without interfering with each other. Those passengers should also be ordered so as to eliminate the need to pass by anyone either in the aisle or in the rows. In other words, it is better to make passenger boarding a parallel process where multiple actions occur simultaneously, instead of a serial process.”

To satisfy my need for excruciating detail and evidence, I read through Steffen’s 2008 research article in the Journal of Air Transport Management. As per the abstract, “Using a Markov Chain Monte Carlo optimization algorithm and a computer simulation, the passenger ordering that minimizes the time required to board an airplane is found.” (I will admit, after reading the abstract, I almost gave up…but then I kept at it).

Here’s a PDF of the research paper: Jason H. Steffen – “Optimal boarding method for airline passengers” – Air Transport Management, 2008

In his research, which is best summarized by his Washington Post article, Steffen builds on the same optimization technique used to answer the famous “Traveling Salesman” problem: given a set of cities and the distances between them, what is the shortest possible route to visit each city exactly once and return to the original city? The resulting Steffen method features an airline boarding procedure whereby “adjacent passengers in line will be seated two rows apart from each other. The first wave of passengers would be, in order, 30A, 28A, 26A, 24A, and so on, starting from the back. (For a typical airplane there would be 12 such waves, one for each seat in a row and for odd and even rows.)”

Steffen conducted a field test of his proposed method versus others, using a mock Boeing 757 fuselage with one aisle, 12 rows of six seats, and 72 passengers. The experimental results show that the Steffen method outperforms current industry practices, with a 2x time advantage over back-to-front boarding, and a 20-30% improvement on random boarding order. Depending on the specific aircraft used, “the optimal boarding strategy may reduce the time required to board an airplane by a factor of four or more.”

Here’s a PDF of the experimental results: Jason H. Steffen – “Experimental test of airplane boarding methods” – Air Transport Management, 2011

The main problem to implementing the Steffen method is getting passengers to line up exactly as prescribed. In Steffen’s own view, the primary benefit of using his method is that “it allows an airline to measure how much room there is for improvement and identifies where that improvement is to be found.” I think that airlines would consider this proposed boarding process (or something similar) if it were framed in the context of the potential cost savings from its use. I recall watching a documentary on Emirates’ ground operations at Dubai International Airport (DXB), where the airline has placed digital clocks in front of every plane at the gate to ensure precision arrivals, turnarounds, and departures. For every second beyond its scheduled departure that a plane remains at the gate, an airline loses thousands of dollars caused by late fines and delays to other aircraft waiting to park. This is a business dominated by intense negotiations between airlines and airports over the use of parking gates and arrival/departure windows. I’m sure that airlines would react positively to the notion of speeding up the boarding and/or turnaround process, if the ultimate result were to allow them to handle several additional flights per day or save on airport gate costs. And passengers would be relieved from the presently aggravating experience of boarding a plane. Sounds like a win-win for all parties involved.

The Sikorsky S-76C++

Idea Analysis Blog 2: Industry Trends and Substitute Products

In the last few days, I have focused on making my idea network more diverse. I established contact with StartupBoeing, a division of Boeing that exists to help aviation entrepreneurs navigate and adapt to the various challenges in the industry. By talking to members of their team, I hope to develop a credible database of information about the competitive landscape surrounding my idea.

This week, I focused on getting an overview of industry trends and substitute products. I confirmed my initial suspicion about this venture: it is really difficult to start an airline in the current market. Although the global aviation industry is picking up, with passenger growth expected to average 5%/year for the next 20 years, airlines have begun devoting their entire efforts towards capacity utilization. In order to cope with a variety of dismal macroeconomic factors, especially unprecedented triple-digit oil prices, today’s airlines require incredibly efficient, low-cost planes in order to maintain and boost profits. Companies like Boeing and Airbus have met this demand by creating new fuel-efficient fixed-wing aircraft; however, this presents a problem in my business idea.

Up until now, I had assumed that there would be a low-cost, high-capacity helicopter that would be more efficient to operate than the closest substitute product (a fixed-wing aircraft). I now had to test my hypothesis – and I quickly found many articles online claiming helicopters to be far less fuel-efficient than fixed-wing planes. Most of these claims were unsubstantiated, so I did my own analysis, comparing several high-capacity helicopters to the substitute fixed-wing aircraft that are popular on short-haul routes:

MODEL Aircraft or Helicopter? Passenger Capacity (1-class configuration) Flight Range (nautical miles) Fuel Cost ($ per nautical mile) Average Aircraft List Price ($ millions)
Boeing 737-800 Aircraft 189 3,115 6.38 90.5
Airbus A320 Aircraft 180 3,300 7.66 91.5
Sikorsky S-76C++ Helicopter 12 345 4.89 7.9
Bell Boeing V-22 Osprey Helicopter 24 879 13.75 68.0
Boeing 234 Chinook Helicopter 34 540 22.22 38.55
Kazan MI-38 Helicopter 30 477 13.64 13.0

It is obvious that helicopters have less passenger capacity than their fixed-wing substitutes, and I expected lower flight range as well. However, there are several helicopters that have lower fuel costs per nautical mile than regular aircraft, such as the Sikorsky S-76C++. I did a back-of-the-envelope calculation and found the fuel cost per passenger to be similar between the 737 and the S-76C++. This posed a major question: given these numbers, how can I expect to compete on a cost basis with fixed-wing airlines?

The answer is simple, but presents yet another assumption in my idea. Traditional low-cost airlines don’t pass on 100% of fuel costs to passengers in the form of fare increases; instead, they find alternative revenue streams to make up this burden (e.g. charging for on-board services). I will need to hunt for these additional revenue streams if this idea is to be somewhat viable. Looking for niche cost savings is also a good idea – for instance, landing fees at helipads are much lower than those charged to regular airlines at major airports. There is a lot more research to be done, with several new hypotheses to be tested and calculations to be made.