Bike valuation
How Cyclesite values bikes
The data sources, model and refresh cadence behind every Cyclesite valuation. We're honest about what we know, what we estimate, and what we cannot price.
What feeds a valuation
Every Cyclesite valuation blends three independent signals. None of them is enough on its own; together they give a fair estimate even for bikes with limited direct sales history.
Sold prices: completed transactions across the UK used-bike market. We weight these heaviest because they reflect what buyers actually paid, not what sellers hoped to get.
Live UK listings: current asking prices on Cyclesite and across the broader UK market. These keep the model in step with today's demand rather than last year's comps.
Catalogue and RRP intelligence: original manufacturer pricing, frame material, groupset tier and category. Used as the anchor for newer models where direct sales history is still thin.
How the model produces a number
For each query we look up matching sold and live records by brand, model, year and condition. If we have at least three trustworthy comparables we lean on those directly. Below that threshold we fall back to a depreciation curve calibrated per category and condition tier.
Condition is applied as a multiplier rather than a switch — the gap between Good and Excellent on a £2,500 road bike is not the same as on a £600 hybrid, and the model accounts for that.
The result you see is the mid figure. The slider's low and high band reflect the distribution of comparable sales we found, not an arbitrary plus-or-minus percentage.
How often the data updates
Sold prices land continuously as new sales complete. Live listings refresh on a daily cycle. The depreciation model is retrained weekly against the most recent data, so the fallback figure is never more than seven days behind the market.
If you value the same bike twice in a week you may see a different figure. That is the model picking up new sales, not a bug.
How we define each condition tier
Excellent: minimal wear, no damage, ridden under approximately 500 miles. All original parts in working order. Service history complete or recently serviced.
Good: normal everyday use, light cosmetic marks, drivetrain still well within service life. A buyer expects to ride the bike home without immediate maintenance.
Fair: visible wear, may need a service or one or two replacement parts. Drivetrain or brakes likely partway through their second cycle.
Poor: significant wear, mechanical issues, or damage that affects use. Sellers usually accept a steep discount to move the bike quickly.
What the headline figure does not include
Aftermarket upgrades. A wheel swap from alloy to carbon can add hundreds of pounds; the model only sees the original specification, so factor upgrades in yourself.
Service history above and beyond the condition baseline. A documented full service in the last 90 days is worth a small premium that we do not bake in.
Bespoke or limited-edition colourways. Some specs hold their value better in the secondary market; the model treats them as the standard variant.
Crash damage that is not visible in stock photography. We do not see your bike, so cosmetic blemishes belong in the listing description and the price.
Honest limitations
We are stronger on bikes built since 2015 than vintage frames. If you have a 1980s Italian steel frame, treat the figure as a starting point and lean on classic-bike forums.
We are stronger on UK marketplace data than imported bikes. A US import or grey-market frame may price differently here.
We do not value parts on their own. The model is for complete bikes.
When a bike has fewer than three trustworthy comparables, the slider's range widens to reflect that uncertainty. A wide range is the model telling you to expect more variation in offers, not a softer figure.
How seller feedback shapes the model
Below every valuation we ask sellers to drag a marker to where they think their bike should sit. When at least 30 sellers of the same brand-model agree on a tighter or looser figure, that signal feeds back into the model on the next training run.
Your input is anonymised and aggregated; we never expose an individual seller's price expectation.
Got a bike to value?
Start a valuation