Why two apps show different elevation gain for the same hike

The short answer

Elevation gain is not a single measurement. It is a running total of every upward step in a sequence of altitude readings, and altitude itself is the least precise number a GPS device produces — noisier than latitude and longitude by a wide margin. Two apps recording, or two services processing, the same walk can start from different altitude readings, decide differently which small wobbles count as "real" climbing, and finish with totals that are both defensible and quite far apart.

Altitude is measured worse than position, by design

A GPS receiver works out its position by comparing the timing of signals from several satellites spread across the sky. That geometry is far better at pinning down where you are horizontally than how high up you are, because the satellites useful for a horizontal fix are spread around the horizon, while the ones that would help most with height tend to sit closer to directly overhead and are fewer in number at any given moment. The practical result is that a phone's raw GPS altitude reading wanders by several times more than its horizontal position does, even standing still. Any total built by adding up changes in that reading inherits all of that wander.

Two different instruments, two different kinds of error

Some devices lean on GPS altitude alone. Others also read a barometric pressure sensor, which infers height from air pressure rather than satellites and tends to track short-term changes in height more smoothly than GPS does. The trade-off is that barometric readings are relative, not absolute — they need a starting elevation to calibrate against, and ordinary weather moving through over the course of a few hours changes air pressure at a fixed spot, which a barometric altimeter cannot tell apart from actually climbing. A calm day and a day with a passing weather system can produce different elevation gain totals for the identical route, from the identical device, for this reason alone.

How much wobble gets counted as "climbing"

Even with one clean stream of altitude readings, there is a choice left to make: how much of a small up-and-down counts toward the total. Add up every recorded increase with no filtering at all, and ordinary sensor noise — a metre up, half a metre down, repeated thousands of times over a long hike — inflates the total well past what was actually climbed, because a jittery reading rarely cancels itself out over a sum. Filter too aggressively, smoothing the trace or ignoring anything under a fairly high threshold, and real short climbs on undulating ground get discarded along with the noise. Different apps sit at different points on that trade-off, and there is no single correct setting — only a choice about how much noise to accept against how much real climbing to risk losing.

Some services replace your altitude entirely

A number of platforms do not use the altitude your device recorded at all. Instead, they take the latitude and longitude of each point and look up an elevation for that spot from a separate terrain database, then compute gain from those looked-up values. This "elevation correction" tends to produce smoother, more repeatable totals, because a terrain database does not jitter the way a live sensor does — but it is describing the shape of the ground at your coordinates, not what your device measured as you walked over it, and it can disagree with a raw-altitude total by a meaningful margin on the same route.

A few other things that quietly change the number

• Whether an app measures gain (climbing only) or the difference between highest and lowest point — these are not the same figure and get confused for each other often.
• Terrain databases are not all built at the same resolution, so two services doing elevation correction on the same coordinates can still disagree with each other, not just with a raw GPS total.
• A route recorded near a lot of tall buildings or under thick tree cover degrades the GPS altitude reading further, on top of the ordinary vertical noise every reading already carries.
• A device that has not been outdoors recently may start a barometric altimeter from a stale calibration point, shifting an entire hike's readings up or down by a roughly constant amount.

What to actually do about it

There is no version of "the true elevation gain" waiting to be uncovered by trying harder — altitude is inherently the noisiest reading involved, and every app makes its own reasonable choices about filtering it and, sometimes, about whether to trust it at all over a terrain database. Chasing an exact match with a friend's app or a trailhead sign is chasing a number nobody was really measuring the same way you were. What holds up is comparing a route against your own past routes recorded in the same app, since at least the method behind the number stays constant even when the ground does not.

TrailCam records elevation gain alongside distance, moving time and pace for every route, and shows the elevation profile underneath a replayed hike so you can see where the climbing actually happened, not just the total at the end.