By: Steve Tengler
What does “ineffable” mean? Viewers of the apocalyptic Amazon Prime show “Good Omens” frequently hear that word during exchanges between Aziraphale and Crowley, the angel and demon played by Michael Sheen and David Tennant respectively. “God does not play dice with the universe; He plays an ineffable game of His own devising.” For those viewers who bothered to Google the vocabulary, they found ineffable to mean “too great or extreme to be expressed or described.”
Interestingly, that adjective (and several others) describes many projects in automotive these days: so large the full requirements are … well … ineffable. Jaguar Land Rover estimates autonomous vehicles will require one billion lines of code, which is seven times the amount of code in the 2016 Ford F-Series, and this does not include the expansive software required for cloud-based services across the end-to-end user experience ranging from retail sales to prognostic service systems to advisor-assisted navigation.
How that plays out between the manufacturer and the Supply Base varies depending upon the customer and project. Some projects follow the path of the original Cadillac User Experience (CUE), where the team produced 3,000+ pages of specifications over 1-2 years with supporting personas, digital simulations and story rooms filled with Post-It Notes®. In the end, no one truly knows such a specification, the coding shall be delayed, the user experience will lag the marketplace and multiple executives will either be demoted or shown the curb. On many other projects, the manufacturer sources a late, incomplete Statement of Work, attempts to shoehorn unreasonable efforts into a starved timing plan and then floods the suppliers with a deluge of Change Requests to flesh-out the missing details. The inevitable ending is a series of fire drills near Start of Production (SOP).
The seemingly obvious solution to these define-it-all-and-late projects is to just timebox the development, prioritize the work and fill-in the details later. “We want three years of development by 30 coders.” However, the near-commodity margins of automotive cause Purchasing to scrutinize and negotiate on anything feasibly devolved into an hourly rate, so timeboxed concessions force the suppliers into using offshore, low-cost resources. Imagining and producing a creative, differentiating set of features is difficult enough with an experienced, centralized staff, but becomes frighteningly difficult with dispersed, low-cost contractors.
So how does a company avoid both ineffable and f’able?
The Better Mousetrap
In 2012, an auto manufacturer — who shall remain nameless — sourced three vendors for an innovative, “dealership reputation management” project with an aggressive timeline and only high-level requirements. No excruciating, detailed-requirement documents. No prolonged, negotiating sessions with protracted counter-proposals and agonizing concessions. Just three trusted vendors and “go.”
“Based off of the input from our team, the manufacturer put together the original project’s rough requirements with key elements and milestone timing,” says Erica Sietsma, the COO of Digital Air Strike (DAS), one of three aforementioned suppliers. “Each of the suppliers helped troubleshoot the necessary, interface requirements so the overall system could work, but most of a given supplier’s work was building out customer-centric stories, developing the designs, running them past the manufacturer, and improving their proprietary offering.” In effect, the scheme was akin to Guy Fieri’s cooking show “Guy’s Grocery Games” where the contestants are given a few ingredients, a thinly-defined theme and a deadline, but no exact recipe.
But here was the genius of the project: The end customer could select from any of the three designs based upon the better user experience and, therein, reward the supplier who competed the best. If customers selected Digital Air Strike’s design, they gained more revenue. “They challenged us to provide the best possible product, and essentially only paid for the products that were used,” stated Sietsma. The open-ended, agile nature of the project allowed for a quicker start and greater competition. “The manufacturer did eventually lock the price point, but that didn’t stop the suppliers from adding more features and improving existing ones in order to win additional business,” Sietsma said.
Eight years later, this project has been rolled out to several thousand dealers and continues to thrive via small improvements on the stable platform.
Less is more.
“Some people would look at this and say, ‘There could have been a lot of failure points since it wasn’t that documented.’ That’s scary, right?” Sietsma smirks and tips her head. “But this is one of our best, long-standing projects. We’ve seen many others come and go. I truly believe having the openness to rethink the project design helped it be as successful as it has been. The Agile Development structure allowed for our customers to win, and their customers to win. That’s what we all want, right? We continue to use it as a case study for our other clients.”
Could this happen for in-vehicle technology? Certainly. The most obvious case would be an end-customer, showroom competition for user experience, e.g. the centerstack Human-Machine Interface (HMI) systems from Apple, Google or Amazon. That would only work, however, for a small fraction of systems since neither the dealers nor the end customers have the patience and time to handle a smorgasbord of menu options. But the competition need not be just for tangible or branded experiences. It could be a blind Pepsi-vs.-Coke test where JD Power or Consumer Reports scores linked to a specific vehicle reveal which suppliers win additional, future business. It could be invisible user experiences where quality data reigns supreme. By enabling an ongoing competition, the end is NOT nigh.
And then, maybe just like “Good Omens,” the written plan is only the beginning of the ineffable plan.