Just pulled the trigger on a Specialized Allez (61 frame) road bike, since I am getting back into riding. While the Allez does have a carbon-fiber fork and aluminum frame, I think the wheels could be lighter and question other parts that could be replaced. So I looked at upgrades and noticed a Trek Emonda (~$3k) was not bad. Then looked at some Domanes, and there was an "endurance" for (~$4k) and a Domane triathlon for (~$6k). The endurance Domane was very light and had wider tires, but the triathlon Domane had skinny wheels that were carbon-fiber. Unfortunately, for my taste I thought that it would be better to have either a hybrid with the triathlon wheels on the endurance bike, or a hybrid with the endurance handle bars, cables, brake handles, on the triathlon bike.

Given the above, it's known that you have to pay more for lighter parts, hence, the price-per-pound goes up. Thinking about a plot of price vs. pound for a specific part, a part that gets expensive quickly with decreasing weight will have a lower slope (price/pound) than a part that sheds a lot of weight but costs less per pound --> greater slope.

So as a rule-of-thumb, which parts typically have a greater slope which allow you ditch more weight with less cost? What would be the order of lowest slope per part for wheels, brakes, crank, derailleurs, neck, seat post, cables?


Slope is actually negative for a plot of price per pound. Using the attached cartoon, a strong negative slope implies that for each additional pound, you pay maybe $2800 less, hence slope is -$2800 / per increasing pound -- or, for each decreasing pound you pay $2800 more. Ideal part would be as each pound is added, you pay $300 less and not $2800 less, same as, each pound lost costs $300 more - Utopia!

(slope means change in price divided by change in weight between 2 or more bikes. The price of a single bike, like $15000 that weighs 15 pounds does not give you slope. Slope requires comparison via differentials with one or more bikes that costs $1000 and weighs 20 pounds).

enter image description here

  • 3
    I think that if you tried to plot price vs weight for a bunch of parts in a given category, you’d wind up with a scatterplot that doesn’t regress to a line in a meaningful way.
    – Adam Rice
    Nov 23, 2021 at 23:40
  • 5
    IMHO, the component that offers the best $/gram savings is the rider. Very often these weight savings come with a free power increase.
    – mattnz
    Nov 24, 2021 at 3:43
  • 4
    Guuuys, please stop telling people that they should lose body fat or that bike weight is not important whenever bike weight comes up! Some people can’t (reasonably) lose body weight and for some people 1 or 2% of weight and speed improvement is worth the money.
    – Michael
    Nov 24, 2021 at 8:33
  • 3
    General comment: if you truly want to play the weight weenie game, consider visiting the Weight Weenies forum (weightweenies.com). I don't play this game myself for reasons RChung outlined, but nobody will stop you.
    – Weiwen Ng
    Nov 24, 2021 at 18:38
  • 4
    Here in the UK, all parts have an equal slope of £1 per pound.
    – PLL
    Nov 25, 2021 at 9:46

5 Answers 5


The full answer is long and complicated but the short answer is almost surely 1) the proper air pressure at zero cost (meaning, zero slope in terms of price/kg); followed by 2) latex tubes (at a price premium over butyl tubes of just a few dollars; and 3) the proper tires.

The short answer is that rolling resistance can be translated into an "equivalent mass," and since rolling resistance applies to the entirety of mass combining both rider and bike, even a small improvement in the coefficient of rolling resistance multiplied by total mass produces an "equivalent mass" reduction greater than a direct reduction in mass.

The attached figure gives an example of how even small changes in coefficient of rolling resistance can have large overall effects, and is the basis for the aphorism, "if you're a weight weenie you should definitely also be a Crr weenie." For a fuller response (though still not completely full response, which requires some mathematical reasoning), you can watch this Youtube video:

Crr vs equivalent mass by road gradient

  • 1
    Tyres with good Crr also tend to be light, as extra material makes tyres less supple - and tyres are cheap, especially if you upgrade when replacing them anyway. I'm not a racer and take my road bike on rough stuff so run bigger tougher tyres than the OP probably will, but switching from say Marathon Supreme to GP5000 at 32mm will save you 85g/wheel for £5/wheel (probably less if you really shop around)
    – Chris H
    Nov 24, 2021 at 8:56
  • And if counting in air resistance, people optimising for speed without participating in regulated races may want to consider a recumbent.
    – gerrit
    Nov 24, 2021 at 14:11
  • 4
    Hey, air pressure isn't free. By my calculations, inflating a typical 700C tire from atmospheric to 70psi will cost you around 10 cents in white bread.
    – MaplePanda
    Nov 24, 2021 at 19:26
  • The cyclist equations are very "textbook" and not very robust, since robustness means a system performs well in a variety of environments. During Tour De France, with such a large variation in environments and road surfaces, I think steady-state velodrome-based rules about tire tread-type don't hold. The laws of physics (cyclist equations) still apply to Tour De France, but the jaggedness in terrain differences "throws sand in the gearbox" for your models. I'm not saying tread types in the velodrome need to be different from Tour De France, but rather the environment is very jumpy.
    – wjktrs
    Nov 25, 2021 at 2:05
  • @ChrisH You're right, good tires can be light(er), and since tires rotate, weight savings are magnified by (roughly) 1.5x to 2x, so a 1g savings at the tire is worth around 1.5 to 2g of weight elsewhere.
    – R. Chung
    Nov 25, 2021 at 6:48

Do also consider durability - it is a common misconception that more expensive things last longer, and while that may be true comparing mid-range to cheap stuff, the top-end parts tend to have a shorter effective lifespan than mid-range parts.

The classic cycling example is Shimano Dura-Ace, built to be lightweight for high-end racing, compared to Ultegra which is the same shape but made of more durable parts, and therefore weighs more while costing less.

The ultimate example of that is motorsport engines. A Formula 1 engine that could do double the length of a race is overbuilt and can be lightened. I'd much rather have parts that survive.

Weight isn't everything, but everything is a compromise.

  • 1
    Yes, there's no easy button, everything's complex. Probably a lot of non-linearity.
    – wjktrs
    Nov 24, 2021 at 0:37
  • 2
    Mmm…some parts go the opposite way. DA chains are better heat-treated than Ultegra. DA pedals have smoother bearings and better seals. DA bottom brackets are better sealed (but have smaller bearings, so you pick your preferred evil). And F1 teams have 3 engines max per season (or face heavy penalties), so the one-race-then-engine-go-boom trope is false.
    – MaplePanda
    Nov 24, 2021 at 1:01
  • 3
    @MaplePanda I am not sure if DA chains have better heat treatment. They definitely have low-friction coatings on more parts than Ultegra, and Shimano explicitly state this. From listening to a podcast (Josh Poertner, Marginal Gains) that DA chains may have better shaping on their plates than Ultegra. Zero Friction Cycling tested chain life in a Cyclingtips article, and they believe that DA chains may have slightly shorter average lifespan than Ultegra (but no clear relationship for Campy Chorus vs Record, and may be the reverse for SRAM Force vs Red) cyclingtips.com/2019/12/the-best-bi
    – Weiwen Ng
    Nov 24, 2021 at 1:54
  • 5
    The argument of DA specifics is detracting from the point of the answer, that some point, durability is sacrificed to weight, so comparing weight/$ is ignoring another factor. If doing a cost/benefit calculation, it cannot be ignored that durability may be sacrificed.
    – mattnz
    Nov 24, 2021 at 4:24
  • 3
    @MaplePanda in F1 it's more like a complete overhaul every race than discard, but the idea still holds, and that overhaul will use a lot of consumables (that wouldn't be consumable in a normal engine) I'm sure. The bike equivalent to that might be using oil instead of grease in bearings, and sprockets so light that the most-used ones only last a few hundred km
    – Chris H
    Nov 24, 2021 at 8:47

I think there is no easy answer. Which part you should upgrade first very much depends on what’s currently on the bike. But keep in mind, if there was an easy and cheap option the bike manufacturer would have probably picked it already.

Many ~2000€ road bikes come with a relatively cheap and heavy wheelset around 1750g. For 850€ you can get a 1230g wheelset reducing your weight by more than half a kilogram. However this lighter wheelset is probably more fragile. With wheelsets it often also makes sense to improve aerodynamics at a slight cost in weight. So maybe it would have made sense to get a slightly heavier but more aero wheelset.

If you don’t need much saddle padding you can get a more or less pure carbon saddle for ~170€ which will weigh less than 100g compared to a normal saddle at ~250g for ~70€.

If the bike comes with Shimano Ultegra it doesn’t really make sense to upgrade to Dura Ace for the weight saving alone. The biggest Ultegra->DuraAce weight saving is in the cassette, which is a wear part.

In the end there is not a single component you can upgrade (on its own) to reduce bike weight a lot. After all, most components by themselves weigh much less than 1kg. So the potential for weight reduction is limited.

https://www.bikeradar.com/news/shimano-ultegra-di2-r8100/ Shimano Ultegra R8100 vs Ultegra R8000 vs Dura-Ace 9200 vs Dura-Ace 9100 weight comparison

  • 1
    Why do the "shifters and brakes" have a battery?
    – gerrit
    Nov 24, 2021 at 14:14
  • 2
    @gerrit: Because the new Ultegra and DuraAce groupsets have electronic shifting only.
    – Michael
    Nov 24, 2021 at 14:19
  • The saddle and wheel options are interesting. Before I was born my dad used to ride to ride to TTs carrying his fast wheels because those flimsy light wheels (with tubulars) weren't to be risked for mere transport. So a light wheelset for races only has long been a reasonable thing, and, as in the discussion under Criggie's answer, it doesn't have to last very long. In the case of a super-light saddle, it's you, not the saddle, that only has to survive the race, but if you swap for training, you'd better get both positions spot on
    – Chris H
    Nov 25, 2021 at 13:28

You could also abandon some parts (like the bar tape), making it a replacement that cost less than nothing but still provides weight savings. Other than that, though, ride it for a while. Latex tubes are much lighter for the price, but don't hold air pressure as well. You might choose to replace some parts for reasons other than weight (like going to a compact crankset will help with both weight and climbing, but be bad for maximum speed).


Having been part of a lightweight vehicle competition in college, the answer to reducing mass is often to remove anything unnecessary for work and safety and to replace any part you can with zip ties, which are lightweight, durable, and cheap the the point of being nearly free!

Further, as noted in other answers, you'll get more performance on very flat and smooth roads ("ideal conditions"?) by reducing the rolling resistance, but this will make your bicycle perform worse or more dangerously for less-smooth terrains

For more performance per price (low slope to high)

  • better tyres (treaded explicitly for the terrain and water removal - warning unsafe for smoother terrain significantly overinflated to 2-3x mfg-rated)
  • better bearings everywhere you can (at least Timken offerings)
  • better wheels (warning: lighter wheels will break excitingly under shear stresses)
  • better frame (maybe 500USD to comically expensive, but spend as much as you're able to and also have it fitted by an expert - the better the frame fits you, the better you can use your bike, and you'll get far more out of a cheap, but well-fitted frame/seat than an unfitted performance pair)

However, to directly answer your question:

What would be the order of lowest slope per part for wheels, brakes, crank, derailleurs, neck, seat post, cables?

  • seat post (simply remove the seat and post for a great negative slope)
  • brakes (if you have disc brakes, converting to squeeze brakes will be significantly cheaper for your negative slope, but also give much worse performance - it's unsafe, but you can also entirely remove the front system)
  • cables (cables are so cheap, but you can remove some entirely or convert them to nylon string to get more negative slope - may need frequent adjustment)
  • crank (changing to a cheaper, arguably-worse crank parts will save you a little weight here for a negative slope, but they will wear much more quickly - as a joke entry, you could have a friend with a 3d printer make yours, but it likely won't last a mile)
  • derailleurs (convert to a Sturmey-Archer hub)
  • neck (center / irrelevant)
  • wheels (here you can spend as much as you like and see good performance with better wheels to a point)
  • frame (added for effect: here is likely the most expensive part you could possibly replace, coincidentally the largest part, and will almost-certainly have the worst weight savings per price due to the extreme curve after a 1000USD frame)

You could also consider a unicycle, disposing of all or at least half of most parts (beware: this will be slower under non-circus conditions!)

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.