Table of Contents
If ever there were an experienced teacher or wise magi in the realm of fast food, one would think it would be McDonald’s. They have dominated the space for decades and have a brand value twenty times larger than Burger King. That said, recent news from under the golden arches (versus the golden crown) has been less than stellar: a Q2 earnings report in the midst of a global pandemic showing McDonald’s sales fell 30% from the same time a year ago, a resurfaced 24-year old hamburger that supposedly didn’t rot in a closet, a lawsuit with its former CEO over multiple allegations, and now a $1B lawsuit for discrimination at franchises during the Black Lives Matter explosion within the U.S.
So maybe it’s worth another look at what Burger King might teach us right now.
For those who have gotten out of their cars and walked into both establishments, the difference in waiting, ordering, paying and receiving the food is significant. At McDonald’s, the customer must visually select the seemingly opportune line upfront and then stand in that line for better or worse. After committing one’s self, an event inevitably occurs: the ten children of the parent in front return from the restroom (and, of course, are waffling on today’s menu choice), the register malfunctions in a fashion never seen by the neophyte behind the counter, the elderly lady paying in cash cannot find the exact change in her purse, or the befuddled cashier flubs up something in a previous order and is juggling the logistics of multiple ongoing transactions. Therein, the selected line seemingly never is the fastest.
Burger King, though, takes advantage of Queuing Theory, which got its start over one hundred years ago in Denmark and consists of mathematical modeling of wait times in various scenarios. The serpentine line with multiple cashiers is the most efficient queuing system since the occurrence of an impediment does not stop the entire line; only one cashier must tackle that perplexing issue while the others plow through the remaining effort. Additionally, on the backside of the process, a server is able to jump between orders and gather the components of multiple meals based upon availability. When every part has been assembled, that meal is delivered to the customer regardless of the original chronology of ordering. With three registers, a Burger King serpentine line is three times faster than three standalone McDonald’s cashiers. Period. The equations haven’t changed in those hundred years regardless of the wacky offering each month (e.g. bacon sundaes??).
So how does any of this have to do with automotive? The answer: engineering resources. Let’s compare the automotive’s equivalent to McDonald’s staffing and running projects versus where some in the industry are headed after learning their lesson (Burger King).
Automotive’s Equivalent of McDonald’s Lines
Let’s first explain how engineering is frequently staffed. Organizations build what’s known as “functional silos”, which are multiple, centralized expertise(s) in a given area of knowledge. Consider these like the individual registers at McDonald’s: there’s a singular leader (“supervisor”) who manages the work for some core group (e.g. Algorithms, Validation), [s]he receives the work orders and “urgent issues” as they arise and dole them out to the team. This supposedly allows for coordinated, organizational direction for a given process, deep training for areas needing extreme knowledge, and the sharing of gurus in low supply. “[Functional] owners [efficiently] develop the mission and vision, and set objectives for managers and employees to follow when achieving these goals,” explains Osmond Vitez. And such practices become even more prevalent when high-demand-yet-low-supply labor is easier to find – or likely cheaper – in various regions of the world, e.g. software architects in one country and software developers in another country. So suppliers eventually have core teams for systems development near the customer, platform development near the global headquarters and software development near the glut of availability; many times separated by multiple time zones.
Unfortunately, this is the bane of the automotive industry (and likely other industries like aerospace) since this means any given task becomes the equivalent of that customer at McDonalds. It cannot go to whichever cashier becomes available first. It must await the specific group’s readiness. If some impediment arises before it, too bad. It must wait. And wait. And, yes, sometimes the equivalent of the wayward child arises: a major defect on another project has shutdown a manufacturing plant, which signals “all hands on deck”, thereby pausing the queue of work behind it.
What exacerbates the issue is the serial stack-up of the functional silos. Imagine yearning for a combo meal at McDonald’s where burgers, fries and drinks were three separate lines. And, yes, each time there’s inevitably some unforeseen hiccups – except of course it’s foreseeable since it happens every time – and wallah, the term “fast food” no longer apply. Anything and everything takes an eternity.
Faced with these delays, the most frequent scenario, though, is either the skipping of tasks or aggressive people short-cutting the lines and causing unmeasurable postponements to tasks farther back in the lines. Thereafter, maybe the development doesn’t meet the commitment (a.k.a. an “Unhappy Meal”), the testing is rushed or done in parallel, or the deadline gets pushed back to a later date. Regardless of the pick-your-own-ending story, customers end up annoyed or quality suffers. There are 30-50 million vehicles recalled every year with the trend line going in the wrong direction, mostly because of projects being rushed to market.
Life as a King
Now imagine the scenario where experts are all assigned to a project-specific team. Tasks could be done by one of multiple employees when they become available. Efficient. If an impediment hops up at Register #1, the serpentine line can leapfrog that employee and seek resolution at Register #2. Agile. Critics of this recommendation will revive the earlier points about centralized control and scarce experts, but this utopia can still be enabled by deployment of said geniuses to projects in duos or trios. If the project-to-guru ration is 1:1, then utilize these matched pairs on two projects together (i.e. 50% allocation). An even better scenario – which many times is tough because of that geographical separation – is training employees to be multi-specialists so they can field many more tasks as they arise; otherwise known as T-shaped since they are a mile deep on a few specialties and a narrowly aware how to do a few other tasks if paired with a different guru.
And tandem-employee design provides additional benefits beyond the efficiency. Studies have shown that paired programming and peer reviews result in fewer mistakes and have a positive Return on Investment with upwards of a $10 benefit for every $1 invested along with 15% fewer defects. Additionally, it allows for improved cross-training due to forced interactions, better resource-backfilling in the event of turnover, better morale due to lower stress, and better communication among the team.
As said best by the street sign outside a Burger King in western Pennsylvania, “Why work for a clown when you can work for a King?”