This two-part entry is actually the second half of my previous post about why there weren’t nearly enough bikes in model year 2012 and why we don’t seem to be able to do anything about the chronic under- or over-supply of product in our parochial little industry.
One of the big reasons we don’t just order more fill-in product in-season and/or hold off on introducing new models until the previous year’s inventory is sold through at retail is the existence of what I call the Model Year Trap.
On the one hand, the planned obsolescence which lies at the root of the whole Model Year concept theoretically drives more sales. But a number of people in the bike business (you know who you are) are starting to publicly question whether revenue lost from the discounting of old models and the forced introduction of new ones ends up being a net loss for the industry, not to mention the consumers who buy and ride our products.
But before we start slicing that particular pie, let’s back up almost a hundred years, all the way 1920s, with stops along the way at the Eisenhower and the Nixon administrations.
Welcome Aboard The Model Year Treadmill. The basic premise of model years—for hard goods, anyway—comes from the automobile industry back in the 1920’s, when the General Motors prez Alfred P Sloane swiped the whole idea of fashion industry seasonality and applied it to cars. Good idea, although it didn’t do much towards preventing the great Depression.
Bicycles, of course, are even more seasonal than cars. So when Schwinn legends Ray Burch (EVP Marketing and himself a product of the Detroit auto industry) and Al Fritz (VP Engineering) first brought the model year concept to their Krate series in 1968, it’s easy to understand how it became such a huge hit.
The initiative was successful partly because it combated a general move away from Fair Trade pricing nationwide, and more specifically because it prevented price erosion for Schwinn and its retailers. Both concerns came at a time when Schwinn’s grueling, multi-decade antitrust battle with the US Department of Justice was finally grinding to a halt and the company was being forced to give up the kind of absolute price controls over relatively long product life cycles it had long enjoyed. (Our current period of relative in-season pricing stability has only been effectively in place since the 2000s.)
Model years worked well for a bunch of other reasons, too. Dealers could be brought together for a single meeting (or in Schwinn’s case, a series of regional meetings) and be shown a whole year’s worth of products at once. Catalogs could be printed just once annually (remember this was back when millions of very expensive catalogs were a both a must-have marketing tool and a huge chunk of budget for bike companies).
Best of all, dealers could be encouraged (some retailers prefer words like “coerced,” “forced,” or even “extorted”) to forecast a entire year’s worth of bicycle sales at once, weather and markets be damned. And of course, if they ended up with a bunch of unsold inventory left over and had to take a margin hit to move those units out the door, well, that was their problem.
Which brings us to the first pincer of the Model Year Trap: we’re a weather-based industry and no one can predict the weather a year ahead of time.
Duh. But weather’s the main reason we end up with not enough inventory at the end of model years when the weather’s been good (like 2012) and discounting excess inventory in years when it’s not (like 2009, when the weather and the economy dealt the entire industry a massive one-two punch). But instead of adjusting our supply chain model to make it more responsive to this kind of seasonal variance, we've done precisely the opposite.
Pre-Season Ordering = Risk-Sharing. Sort Of. Back when Schwinn Ruled The Earth, factories were domestic, which meant bikes could be fabricated much closer to when they would be needed, and that adjusting production to meet seasonal demands was relatively simple. (Schwinn, old-timers tell me, also deliberately under-produced to about 80% of forecast in order to prevent inventory buildup and keep prices stable.)
But beginning in the 1980s with the shift to offshore production, the need evolved to build up large amounts of inventory in order to level-load production at Asian factories (discussed in more detail in my previous post).
Lead times got longer as a function of geography and global parts sourcing, up to four months between order and stateside delivery, which is where they stand today. But at the same time came the notion of the component model year, itself a direct result of the Shimano’s unprecedented technical development program of the same decade. At one point in the early 2000s, product managers had a six-month (!) wait just to get Shimano kits to their factories so they could start building bikes. So 6 months for kits plus a month to assemble and package plus another month to ship and bring bikes through customs and into warehouses meant an effective 8-month lead time.
With delays like these, year-long forecasts from retailers became a necessity to spread the risk associated with suppliers committing to half-a year’s worth of inventory at a time.
