Having gained experience riding both Sora and Ultegra, I'm puzzled that 2x9 already provides such narrow gaps that I'm entirely comfortable riding Sora (it helps that the cassette is 12-25), and that, by far, the biggest difference between the two groupsets is what could be termed their MTTR, or their Mean Time To Repair. (In a cycling context we could refer to the Mean Time To Retune.)

Ultegra FD and RD do not need maintenance, not even after one year of riding, whereas even a perfect Sora FD and RD tune-up will only last 3 months, and then another tune-up is needed.

Am I right in reaching the conclusion that MTTR is one of the major factors distinguishing these two groupsets? It could be, for example, that I was unlucky with one and lucky with the other, or it could be that I rode Sora on slightly gravelly roads (as the beater road bike) and Ultegra on cleaner open roads, contributing to the faster degradation of the tune-up on one compared to the other. Importantly, is 105's MTTR anywhere near Ultegra's?


  • I don't know that anyone has compiled enough data to answer this question definitively. My guess would be that the road conditions you mention (gravel) had more to do with reduced performance than the components themselves. Generally, the main differences between those groupsets are going to be weight and durability. And I would expect the lower durability of Sora to result in catastrophic failure rather than minor degradation that can be fixed through adjustments.
    – jimchristie
    Nov 28, 2023 at 21:22
  • 1
    As an allegory - a racecar or plane needs a lot of tuning and maintenance regularly, whereas a tractor or runabout can just keep running for years without significant effort. I'd see Ultegra on fast fancy bikes where the rider wants top performance all the time, and lower groupsets on more thrash-bikes that just have to work, but nor perfectly.
    – Criggie
    Nov 28, 2023 at 21:52
  • 2
    @jimchristie I've pushed my Sora setup pretty hard, and I think you've got a point, but weight is more significant. After about 45000km I wore out the rear shifter to the point where it wouldn't reliably upshift after a downshift (there's a little lever that doesn't spring back properly, somewhere near my thumb). The rear mech itself got replaced after my chain ate a twig causing the wheel to eat the rear mech, but is still going strong after 20000+km - at worst a little stiffer
    – Chris H
    Nov 28, 2023 at 21:58
  • 2
    @Criggie I appreciate the allegory, but I'm not being that demanding. I just want to shift smoothly with no cage rub. On 2x11 bikes that means having 10 gears on one chainring and 10 on the other. On 2x9 bikes that means 8+8. It's this creep of chain rub that makes the otherwise terrific Sora unpleasant to own.
    – Sam7919
    Nov 28, 2023 at 22:13
  • 1
    I know this question is about the shifting of the systems, but I will add that the derailleur jockey wheels with bearings (Ultegra, XT) are much better, last longer, and are quieter than the jockey wheels with bushings (105, SLX, and lower).
    – Paul H
    Nov 28, 2023 at 22:43

3 Answers 3


No experience with road groupsets, but with MTB ones (I have 3 bikes, one with XT M8100 (shifter and derailleur), one with a mix of Deore/LX/XT (late 2000's), and one that was originally full Acera, but is now a mix of Acera (FD), Deore 5100 (front shifter, rear derailleur) and Deore 4100 (rear shifter)). Acera is the equivalent of Sora, XT the equivalent of Ultegra, the rest is debatable.

If speaking about the "time to retune", I have a similar observation, but there's a clear difference between Acera and Deore, but almost none between Deore and XT. There are also noticeable differences in reactivity between the groupsets: big gap between Acera and Deore, limited after. This can be explained by the cables, that being said: from Deore, cables provided with the shifters are coated. Also, my Acera had some play (derailleur), so it was sometimes necessary to change 2 gears and then one back. The XT is a bit more reactive than the Deore, but that's mostly linked to the shifter design (that has a feature called "instant release" - the actuation happens in one stage of the stroke, while there are two stages with other shifters - noticeable, but not a must have).

By Ultegra, I guess you mean R8000 (mechanical), because if you take the current ones R8100, there's a big difference: R8100 is electronic only. And a minor difference: Hyperglide vs Hyperglide+ (that allows to shift to harder gears in force). Also, another point to consider is the absence of hydraulic brakes in Sora (groupsets technically include the brakes).

Note that in the case of MTB shifters, there are also differences in "behaviours" between Acera, Deore and XT shifters. Front shifters from Deore and up only have one lever (for 2x, while 2x Acera shifters are just 3x shifters with different indicators). For the rear shifters, the differences are the number of speeds you can change in one stroke (+1 hard and easy for XT), and the index shifter that only works in one direction for Acera, while it works in both from Deore and up. Not sure that such difference can be observed with road groupsets.

  • 1
    Now you've set me thinking - I wonder if an old Deore would be better than the Acera 3×9 FD on my MTB, that keeps struggling to get into the little ring. It may even give me more tyre clearance as well.
    – Chris H
    Nov 29, 2023 at 9:35
  • 1
    @ChrisH the trekking bike (the one with Deore/LX/XT) is a 3x9 (Deore FD, LX Shifter), the funbike (that was originally full Acera was 2x9, but is now 2x10 with a Deore 5100 shifter, but still the Acera FD). On the trekking bike, Deore/LX works more reliably that the full Acera of the "original" funbike. There was a noticeable shifting improvement after the upgrade of the shifter of the funbike as well. So my impression is that shifter (and the cable — Deore comes with coated cables) are the most critical parts.
    – Rеnаud
    Nov 29, 2023 at 9:52
  • I've given up on coated cables and gone back to plan stainless. They kept shedding coating into my housings. I'm in the process of weighing up my options for my hardtail - I need a new rear tyre and would like bigger than will currently fit, but the FD is annoying. It feels like the low limit screw screws itself in and suddenly no little ring. I like the range of my triple (6.2:1) too much to spend a fortune going to 1×11 that would max out just under 5:1 - I ride on road to the trails and don't want to lose my high gears
    – Chris H
    Nov 29, 2023 at 10:07
  • 1
    @ChrisH I totally follow you on the range. Each time I look at gravel bikes, I'm turned off by the lack of range on gravel transmissions — the strange mix on the funbike only exists to have a 46/30 front and 11/42 rear (double chainring + wide range cassette is the nicest compromise for usability for me). Note that it can be achieved now with the CUES range, where you could get a 46/32 front and 11/45 rear (in the 6000 series, to limit costs).
    – Rеnаud
    Nov 29, 2023 at 10:38
  • 1
    That's why my gravel bike has 2×11 but I still find it a bit lacking. I'm tempted to swap my 46 chainring for a 48, and go bigger on the cassette for my gravel wheelset. In both cases I'd be pushing the limits of the derailleurs (105, but the chainset is this 46/30 FSA - could be good for MTB). A pure GRX drivetrain seems very limited. CUES does look interesting for new builds, but I'm not an early adopter and would like to see how availability and compatibility settle down.
    – Chris H
    Nov 29, 2023 at 10:45

My experience is with Sora and 105, but is at odds with yours.

I run 3×9 Sora on a tourer and 2×11 105 on a gravel* bike, and have had much more experience with the former. I ride them in similar conditions - clean and dirty roads, and off-road, often in the same ride.

Typically, I might change a rear derailleur cable, tune it again after tens of km to allow for initial settling in, then leave it alone. After that, I might need to adjust it once more after say 3--5000km, but beyond that it means the cable is going and it's time to replace it. That's been my experience with Sora over a few years, and 105 seems very similar this year. The Sora RD may be a little less sharp at times, but generally it needs a good clean rather than adjustment.

One thing to watch out for is that my Sora tourer is QR; if the rear indexing feels off, that probably means the wheel isn't sitting quite right in the dropouts. I have felt the same thing on my thru-axle 105 gravel bike - some grit on the locknut stopped the axle doing up in quite the same way as it should.

At the front, I've found my 105 setup to need more adjusting. The tourer was a problem at first but that was a frame-mounted barrel adjuster that wouldn't stay put. Now it never needs adjusting. The first time I saw the equivalent of a barrel adjuster on a 105 FD was a friend's bike, when it had come out of alignment; mine has done the same as couple of times, and the adjustment is fiddly compared to a barrel adjuster (it should be simpler, but it sticks).

Measuring in units of time is rather pointless, unless you log pedalling hours - things don't degrade like this is the bike shed. Measuring in distance is better but can't cope with the difference in conditions that you note. In the case of shifting components, you may have to consider vibrations, but even on smooth roads, terrain makes a big difference because an area with low but steep hills (as where I live) makes for a lot of gear changes.

* I bought it as a faster endurance road bike than the tourer, and swap between road and gravel wheels as required.


I lack personal experience with the current Sora. My understanding is that the current Tiagra (4700) way outperforms its pricepoint, but it is two steps above Sora.

In general, borrowing the economics concept, we would expect diminishing marginal returns to any input including cost. We can clearly see this in Ultegra vs Dura Ace, 105 vs Ultegra, or the equivalents: the cost is significantly more, but the performance is essentially identical but for half a pound of weight. Conversely, at the lower price points, we would expect that at some point, saving money can cut performance unacceptably.

In the last group test that Cyclingtips did before their parent company folded them into Velo, the reviewers observed that some mechanical disc brake bikes were good, but some had unsafe braking performance. Economizing on cables and housing is one of the things that product managers might do to hit a price point. And we would expect Claris bikes to have cheaper housing and cables than 105 or Ultegra bikes. Thus, I would expect Claris to get out of tune faster than Ultegra, but most of the difference might be explained by poorer cables and housing.

To make cheaper groupsets, you definitely see Shimano economizing on material quality and construction methods, e.g. plastic vs ceramic disc brake pistons, bushings vs bearings for pulley wheels, heavier aluminum alloys in the lower groups, cheaper construction methods overall (e.g. Ultegra and DA now the the Hyperglide+ shift ramps, but 105 omits these due to the manufacturing costs, leaving it with 'just' Hyperglide shifting quality). Tolerances are another area where you might see manufacturers economize. To some extent, material quality might be correlated with tolerances (e.g. it might be harder to hit tight tolerances on lower grade steel or plastic).

For Claris to be more prone to getting out of tune than Ultegra, the materials would need to be weaker in such a way that they might degrade over time, and the tolerances would need to be poor enough to impact the shifting. I would guess that Shimano hits tolerances closely enough, even on Claris. Same with material quality. So, if the durability of Claris falls short of Ultegra, it probably still has sufficient durability to accommodate the sort of riders who might use Claris - new cyclists who buy a bike, then get serious, then upgrade, or occasional cyclists.

And it's worth noting that some of the higher-performance groups may be less durable in some respects. The crankset delamination fiasco was hopefully an outlier, but we can say that if you make things lighter and lighter, you raise the risk of the component failing in general use or of systematic manufacturing flaws. The polymer coated cables found on Ultegra and DA can flake and jam up the housing - I did have a set of Ultegra cables last 3 years, but not everyone gets that experience. The titanium cogs in DA are a clear example, and they wear out fast.

The bottom line is that I think that materials and manufacturing tolerances account for the bulk of the performance and durability difference between Sora and Ultegra as groupsets. Considering them as installed parts of the bike, product spec decisions will make some difference as well, although they can be addressed (although if the bike has through-headset cable routing, that can be expensive).

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.