Cars are terrible products. Not only are they expensive, dangerous, and arguably ruining the planet — worse than all of this is that they break all the time.
Worse, they usually break in stupid ways. I’ve taken my Subaru in for new brakes, sure, but I also took it in that time AC condensation was dripping through the glove compartment, and that time I dropped my keys and the remote buttons wouldn’t unlock the car anymore.
For every dollar Americans spend on new cars this year, they will spend about 46 cents on aftermarket parts to replace pieces of cars that have stopped working.† (That’s not even counting installation costs, or things like tires, which are also expensive but which people know wear down no matter what.)
The aftermarket auto parts industry is a big, steady business — but also, connoisseurs of business quality will be happy to tell you — a pretty mediocre one.
Why? Parts suppliers sweat to make a million complicated little things to sell them at commodity prices, then get pushed around by the OEMs — which they compete with — and by the big wholesale and retail distributors that own the customer base (dealerships, service shops and do-it-yourselfers). Many parts manufacturers — there are thousands, most of them small — get by with a focus on a narrow product category. They get reliable revenue streams some of the time and low margins all of the time.
Dorman Products is an auto parts supplier, but it is not a bad business. In fact, it is a pretty good one. It has a sustainable competitive moat, 40% gross margins, and a track record of 15% annualized shareholder returns since its IPO in 1991. So what’s going on?
Dorman follows a totally different strategy. It begins by taking a few common assumptions and flipping them:
Dorman’s strategy is about new products — literally, about 4,000 of them, every single year. Small-cap investors who stumble across Dorman quickly identify this as the reason Dorman consistently grows faster than the aftermarket as a whole — it has grown revenues at 13% annually in an industry that grows at 3-4% — but it is more than a way to turbocharge growth; it is the only way Dorman exists at all.
Dorman depends existentially on new products. About 20% of revenues at a given time come from products launched less than two years ago. If it stopped innovating, it wouldn’t revert to market-rate growth. It would go bankrupt. Dorman isn’t good enough in any product category to survive in stasis.
The reason to study Dorman as an example of a good business and not a horrendous one is that it has the operational pieces to actually back up its unorthodox strategy of making innovation the core of its offering. The innovation that sustains it is repeatable; its business model is enduring.
One of Dorman’s key insights is that it is possible to predict which parts are going to break. That leaky air conditioner, or that fragile remote key? Dorman probably saw those coming. Months before I came in, hundreds or thousands of other people had probably brought in Subarus with the same problem. If the service technicians noticed a pattern — or if it showed up in repair data — Dorman might well have started working to make that product so I’d have the option of buying it from them. Until then, I’d have no choice but to buy Subaru’s part.
Dorman’s salesforce and product engineers actively tap its end-users — the technicians who are installing replacement parts daily — to systematically identify before anyone else what problems are about to bring loads of customers into the service bay. A fine-tuned marketing machine that includes newsletters, how-to videos, and social media serves a dual purpose that augments this goal: it sells today’s products while connecting Dorman to the people who, in turn, provide ideas for tomorrow’s products.
“Its unique structural advantage lies in its market intelligence as it relates to the failure rate of auto parts,” one of Dorman’s few institutional admirers recently observed. “Specifically, the data collection and the database that it’s amassed over time is really quite unique in the industry. No one has been able to replicate it.”
A recent video for customers provides a sneak-peek of some things you may be buying yourself for Christmas this year: brake line units that corrode, hybrid batteries that fizzle, even little plastic radio knobs that are apparently falling off in epidemic fashion because of some flaw in the original design.
Importantly, the outlook for none of these products is particularly bright. Unit sales will be modest— there are only so many 2008-2011 Mercury Mariners with busted door lock actuators. Within a couple years, competitors will ramp up production of the identical (or nearly identical) part, and prices will fall. In the interim, however, Dorman will have enjoyed a dominant market position in thousands of small markets, and earned generous margins on those early sales.
Dorman is an engineering firm managing a portfolio of products that it creates, sources, markets and distributes. One of the operational choices that enables its strategy is outsourcing all manufacturing. Dorman remains nimble with this “asset-light” model, able to spread itself across disparate product categories and pick and choose from the relevant manufacturing capacity available in the world at any given time (about 75% of parts are sourced in China). While manufacturers do all the work, Dorman retains the rights to the “tooling” developed in the manufacturing process, and to any associated intellectual property. Doing so not only prevents partners from sharing innovations with competitors, but allows Dorman to pick up and move to another manufacturer when, for example, a given product line is to be expanded, and use the existing tooling as a starting point.
From start to finish, Dorman can complete the entire process of getting a product to market in 9 to 18 months, depending on the product’s complexity:
From among tens of thousands of new product ideas, Dorman will choose a subset that promise to be good aftermarket products. It is looking for problematic parts requiring future replacement within niche categories likely to be paid little attention by the OEMs or by competing parts suppliers.
In many cases, it will improve upon the OEM part if it identifies a design flaw that is contributing to the problem — adding a protective coating to a fast-corroding part, for example. In other cases, it redesigns the product to make it easier to install, making service techs happy, loyal, and marginally more likely to check the Dorman catalog the next time they are about to order an OEM part.
Once Dorman has placed an order with a contract manufacturer, that supplier is on the hook for increases in raw input costs and for any quality problems Dorman inspectors discover down the road. (Quality oversight, as with any business outsourcing manufacturing, is a key function for Dorman to perform well.)
Shipments from suppliers are received in warehouses strategically located near those of large customers (e.g. NAPA) in the central U.S. The big customers, in turn, distribute out to their branch networks. Dorman offers a limited warranty and processes returns of defective parts.
Around this point in the process, Dorman begins again to resemble, well, a commodity auto parts business. It must invest in massive amounts of inventory (about 150 days) while customers squeeze it for credit (94 days of receivables), manage stock of 140,000 SKUs, and pick orders from among them. It is the most challenging and logistically nightmarish aspect of Dorman’s business.
The good news is that the growth and high margins provided by Dorman’s innovation machine provide the cash to support these needs (the company’s growth has been almost entirely internally funded). In the end, if auto parts remains a tough business, it’s possible still that Dorman has reverse-engineered a way to make it a somewhat better one.
†While new car sales are cyclical, take for example 2014 U.S. new car and light truck sales of about 16 million at an average price of $33,000 each = $528 billion. For comparison, spending on aftermarket parts alone for such cars was about $246 billion, not including labor and tires.
 Of 4,218 new products introduced in the twelve months ending September 2015, 32% were entirely new products previously unavailable in the aftermarket (i.e. were “OEM-exclusive”). The remaining 68% were extensions to existing product lines—such as an adaptation of an existing SKU to fit a different make or model. Products were mostly in small categories within the broader categories of auto parts: 37% powertrain, 29% auto body, 26% chassis, and 8% hardware.
 Dorman Products investor presentation, Nov. 2015, p. 7
 Investor presentation, p. 9
 Ad Age, “How Auto Parts Manufacturer Dorman Products Revved Up Its E-Mail List,” Feb. 24, 2011. < http://adage.com/article/btob/auto-parts-manufacturer-dorman-products-revved-e-mail-list/283781/>. Dorman’s own website, www.dormanproducts.com, has a modern design and features many customer-friendly engagement tools.
 Steven McBoyle, Royce Associates. Royce Premier Fund: Looking for Enduring Franchises, Sept. 22, 2015. The quality-focused asset manager currently owns about 4% of the company. <https://www.roycefunds.com/insights/2015/09/royce-premier-fund-looking-enduring-franchises>
 Autodata, 2014 Light Vehicle Sales, via The Detroit News, “2014 Car Sales Best In U.S. Since ’06,” Jan. 5, 2015 <http://www.detroitnews.com/story/business/autos/2015/01/05/automakers-report-december-sales/21277199/>
 Kelley Blue Book, via USA Today, “Average New Car Price Zips 2.6% to 33,560,” May 4, 2015 <http://www.usatoday.com/story/money/cars/2015/05/04/new-car-transaction-price-3-kbb-kelley-blue-book/26690191/>
 2015 Auto Care Association Factbook