Integrated Road Bikes Perform Better, but at What Cost?
At moderate fitness, it takes me roughly 27 minutes to ride from my driveway to the gate in Boulder Canyon, and almost exactly another 15 minutes to Lucy’s Overlook. I know the climb intimately: where there are short, steep spots to power over and where I need to shift down and keep my cadence high.
But today I’m almost a minute faster from the house to the gate, and almost 30 seconds faster to the overlook. It’s not because I’ve gained fitness. Rather, today I’ve traded in my usual bike for a Factor Ostro VAM, a stunningly high-performance (and high-priced) race bike optimized for this very outcome. Every surface has been meticulously shaped for aerodynamic efficiency, with airfoil cross sections even on the seatpost, and all shift and brake cables entirely hidden inside the frame and cockpit.
It’s perhaps the furthest current expression of a new trend, called integration, which started in performance road bikes and is now trickling into lower price points and even gravel and mountain bikes. In an attempt to chase ever more marginal gains in stiffness, aerodynamics, and weight savings, designers have started ditching stock handlebars, stems, and other components. Instead, many design their own parts that are integrated with, and often proprietary to, the frameset. “If you’re designing a bike where you don’t have the requirement to use a standard handlebar, stem, or seatpost, you can model it as one piece,” says Graham Shrive, director of engineering at Factor Bicycles. That freedom lets engineers use shapes they otherwise couldn’t. The resulting frames can save weight by shaving material where it’s not needed; or they can have a smoother transition between cockpit and frame for better aerodynamics. But those watts gained and grams lost come at a cost.
First: integration reduces component choice, lessening the consumer’s ability to customize both bike fit and ergonomics. Second: if something does go wrong, an integrated design often defies a simple fix.
Canyon’s new Aeroad is a prime example. It was well received when it launched last fall. But in early March, Alpecin-Fenix pro racer Mathieu van der Poel—Canyon’s highest-profile sponsored athlete—almost crashed when a portion of his Aeroad CF SLX’s carbon fiber handlebar snapped off without apparent cause or warning during a race.
The next day, Canyon issued a worldwide “stop ride” order for owners of the affected Aeroad CFR and CF SLX models. On most bikes, you could simply swap out the handlebar and be back to riding right away. But the CFR and CF SLX feature a special one-piece handlebar/stem called an Aerocockpit for fully internal cable routing. No other handlebars—not even Canyon’s other Aerocockpit designs—are compatible with those frames.
Two weeks after van der Poel’s accident, Canyon delivered the worst news: it would take months to redesign and manufacture a replacement Aerocockpit, during which time owners shouldn’t ride their bikes. This is an example of one of the signal weaknesses of integrated bikes, which is that they’re more likely to be what I call brittle: a problem with one essential but proprietary part can effectively brick an otherwise perfectly functional bike.
Most riders, even enthusiasts, don’t race. The watts and seconds gained from integrated designs matter less than ride quality and riding time, which are hard to get if your bike isn’t functional. So as the industry increasingly adopts integration as a design approach, it’s worth asking: what do we gain, and is it worth it?
How Did We Get Here?
The integration push really got going about 15 years ago when brands started designing frames and carbon-fiber forks as single units with proprietary headset configurations.
But three other developments really pushed the trend forward: internal cable routing, electronic shifting, and then wireless electronic shifting. These design shifts and technologies all came about in close sequence, and they fed off one another in a somewhat circular fashion. Internal cable routing became increasingly necessary once brands started going electronic (electronic wiring looks messy if routed outside the frame, but luckily electronic wires are small, so fully internal configurations were possible). In turn, the hassle of stuffing wires into stems and headtubes led the industry to turn to wireless drivetrains. And wireless drivetrains, combined with carbon fiber’s blank-page design capabilities, gave bike makers even more freedom to accelerate their fully integrated cockpit creations.
So in a long-way-round fashion, the push for integrated bike designs created a problem that the bike industry solved by getting rid of cables almost entirely, thus freeing up even more opportunities for integration.
Bike makers like to say they’re only responding to what sells; if people still wanted metal, rim-brake, nonintegrated performance bikes, they’d make those. There is truth to this assertion. Demand for rim-brake wheels and cable-actuated shifting has plummeted at the high end of the sport. But the full reality is a lot more complex. From what’s advertised to what’s available in shops (in normal times anyway), the industry markets and sells what is new, and most of the design and product-spec decisions aren’t driven by cyclists but by bike brands and their suppliers. Cyclists buy these bikes in no small part because they’re told integrated designs are better, and that’s what there is to buy. But is it—for most, non-racing cyclists—actually better?
What Do We Lose with Integration?
The bike industry tells us these bikes are better because of attributes like weight savings and improved aerodynamics. The aerodynamics story is good marketing, as are the aesthetics: a bike with hidden cables just looks more badass. But the actual performance benefit of hiding cables is negligible: it saves just 12 seconds over 25 miles compared to having all four brake and shifter cables exposed, according to a 2018 test by Specialized.
Even in the case of the Aeroad, Canyon claims the gain from external to fully internal routing is “up to three watts in extreme cases,” which suggests the real difference is even less. (For context, a reasonably fit recreational cyclist can put out 200 to 300 watts at threshold, which may or may not be what Canyon has in mind with “extreme cases.”) I’d argue that we—the industry and consumers—have made this shift mostly out of vanity. More important, we haven’t really accounted for what that choice has cost us.
First, take drivetrains and brakes. Increasingly, bikes are designed to work only with disc brakes and electronic shifting. Those technologies play better with internal routing, even though some riders prefer rim brakes and mechanical-shifting drivetrains for their simplicity and reliability. Component brands have responded: SRAM and, now, Shimano produce their 12-speed groups only in electronic shifting versions, and Shimano’s Dura-Ace and Ultegra are available commercially only with disc brakes. Campagnolo is now the last component brand still designing new cable-activated drivetrains for road bikes.
Likewise, integrated handlebars and stems reduce choice for bike fit and ergonomics. While most bike brands make proprietary bars and stems in decent size ranges, short stem lengths (under 90 millimeters) are especially hard to find, which penalizes smaller riders. Often there’s only one option for stem angle, or handlebar reach and drop. In contrast, independent brands like Zipp and Ritchey offer multiple configurations and models. The complex headset designs and proprietary stems required for fully hidden cables can make it difficult or impossible to add stack height for those who prefer or need a more upright riding position. One-piece bar-stem configurations don’t even allow you to rotate the handlebar to fine-tune your hand position. Some of these brand decisions are small—offering an integrated stem in only five lengths instead of seven. But they add up, and the cumulative effect is a narrowing of options.
It’s important to note what’s forcing these choices. Even big bike companies like Specialized or Shimano don’t have limitless R&D resources. Time and money spent developing aerodynamic tube shapes or wireless electronic shifting have to come at the expense of something. Offering rim- and disc-brake versions of bikes means engineering and manufacturing two different forks to handle the different stresses from each type of braking, for instance. Brands go disc-only because that’s where they’ve decided the majority of the money is.
There’s also the brittleness that I mentioned. Like Canyon, Factor also had a high-profile failure last spring when Israel Start-Up Nation racer Tom van Asbroeck broke his Factor Ostro VAM’s steerer tube at a cobbled race just days before van der Poel’s Aeroad handlebar failure. After investigating, Factor determined (as it described in a detailed and laudably transparent post) that the problem was an improperly manufactured batch of steerer-tube compression plugs that loosened even at proper preload. So ISN mechanics resorted to bonding the steerer plug to the fork and over-torquing the bolt, which caused a stressed riser to break the fork.
Factor ended up replacing plugs on consumers’ bikes, and its in-race failure didn’t get nearly the attention Canyon’s did, but it too highlights the weak point of tightly integrated systems. More recently, Specialized just recalled all of its new Tarmac SL7s for a similar issue with the compression plug. If once is happenstance and twice is a coincidence, then we’re one recall away from a broader question of whether a different compression plug is the fix or if these steerer systems need a fundamental redesign.
Van der Poel’s handlebar breakage was a different issue but may have resulted from a similarly complex group of causes, including that Alpecin-Fenix team mechanics may have overtightened the brake-lever clamp. That clamp is not Shimano’s but Canyon’s own design, required to fit the Aerocockpit’s ovalized cross section. Predictably, it’s taken time to fix. The company tested a new Aerocockpit design with thicker walls and a new, rounder clamp cross section under riders at July’s Tour de France, including during van der Poel’s Stage 2 win. But van der Poel was back on an older version at October’s Paris-Roubaix, and Canyon still has not announced when handlebars will be swapped on consumers’ bikes, even as Aeroad CFR and CF SLX owners are now in their eighth month with a $6,000 to $9,000 aerodynamic sculpture. (Canyon is offering a $1,200 or $1,500 purchase credit to the estimated 1,500 worldwide owners of the faulty models, making for a roughly $2 million hit to the company even before redesign and replacement costs.)
Weeks ago, I contacted Canyon USA’s PR agency to ask about updates on the brand’s replacement timeline, as well as how the episode has affected the company’s philosophy about design integration going forward. My questions were sent to Canyon headquarters in Germany, which never replied despite several attempts to get comment, and owners and prospective buyers haven’t gotten much info either.
That’s a problem. I’ve heard that some Aeroad owners, frustrated with the long wait and sparse communication, just started riding their bikes again with the stock Aerocockpit. If what happened with the Aeroad and Ostro can befall pro team mechanics, then it’s absolutely within the realm of possibility for shop staff and consumers. Van der Poel and van Asbroeck were lucky they didn’t crash; an amateur rider with lesser handling skills might not be so fortunate.
Does Integration Provide a Better Riding Experience?
The short answer is “Yes, but…” For most of us, I’m not sure it’s enough to make the drawbacks worthwhile. I’ve been spending time lately on that Ostro VAM equipped with the new 12-speed Dura-Ace group and C36 wheels. From a performance standpoint, it is superb. I don’t race anymore, but if I did, it would be an ideal choice. Compared to a nonintegrated bike, the Ostro is clearly more aerodynamic. And like other integrated designs I’ve ridden, it manages to pair a sharp, aggressive response under power with a connected-to-the-road ride quality; you’ll still get road feedback, but buzz is somewhat muted. Some traditional bikes struggle to match that level of stiffness without sacrificing ride quality.
I’ve also spent some time testing an Aethos, one of Specialized’s new high-performance, nonintegrated bikes designed around off-the-shelf components. The Aethos is a lightweight road bike that optimizes ride quality rather than aerodynamics. It’s a disc-brake-only, electronic-drivetrain-only bike, but that’s the limit of its restrictions. It has round tubes and uses a standard 27.2-millimeter-diameter seatpost with an external clamp. Cable routing is internal in the frame only, allowing you to choose whatever stem and bar you want. No funky axle spacing or special adapter for the bottom bracket.
It’s an interesting point of comparison to the Ostro. Both are excellent bikes, and they have similarities. But they’re different in important ways, and that matters for the rider. The Ostro is more finicky: during testing, its seatpost was prone to slipping even at proper torque, and I couldn’t easily adjust the handlebar height. But from a performance standpoint, the bike is stiffer under power and more precise on descents. In hard braking for corners, I detected some understeer in the Aethos that I suspect is due to its lack of torsional stiffness compared to the Ostro. But the Aethos is stiff enough for most riders’ purposes. It’s also a super climber whether you’re in or out of the saddle. Perhaps most important given Specialized’s design goals, the ride quality is better than the Ostro’s—actually, better than almost any carbon bike’s I’ve ridden. Not for nothing, several media outlets have deemed one or another Aethos model as their bike of the year (early reviews of the new, gravel-oriented Crux, which has a similar design, are similarly enthusiastic). It feels, most closely, like a lighter version of an excellent steel or titanium bike: carbon for a connoisseur.
Where Is the Trend Going?
There’s a sense that integration may not go much further than it has. “I think it’s going to stall out a bit,” predicts Factor’s Shrive. “You’ve gotta have two wheels and a front and rear brake, so it’s limited how far integration can go.”
I know that there will always be conventional bikes. Small builders especially, who don’t have the capability to engineer integrated designs (or the interest in chasing such marginal performance gains), are always going to rely on outside companies to provide components like stems, handlebars, seatposts, and even forks. But the success of the Aethos, followed by Specialized’s introduction of the Crux, gives me hope that big brands will realize there’s still a solid market for nonintegrated bikes, for riders who want high performance but are focused more on the riding experience than chasing podiums or Strava PRs.
For riders who want to wring every watt and second out of their performances, who rely mostly on professional mechanics for doing much more than swapping a chain or brake pads, who prize an aggressive position on the bike, and for whom compatibility is not a concern because they buy new bikes every three to five years, an integrated bike is the way to go. A model like the Ostro is among the very best of that genre. But a nonintegrated model like the Aethos is probably a better choice if you have a nonstandard fit or need a bit more height on that stem spacer stack; if you hold on to bikes for a long time and want maximum flexibility for replacement parts; or if you simply want the opportunity to customize and outfit your bike the way you like. I know it’s not as aerodynamic—and maybe less stiff under power. But it rides beautifully, and if a component doesn’t work for your fit, or it slips, or even breaks, you can just buy a new one.