Export only giving run times to nearest minute

Kind of a tangent, but I don't like the way Garmin exports CSV for Connect in general. Typically they use some non-normalized format which is ok for humans to look at, but bad if you want to do any kind of automated data processing. But if I just want to *look* at the data, it's already there in Connect. Usually CSV exports are used for further data processing.

For example, if you export laps/intervals from an activity, you get lap times which look like this:

4:30.5

4:45

4:25

This looks ok to a human, who can unambiguously interpret these numbers as "4 minutes and 30.5 seconds", "4 minutes and 45 seconds", and "4 minutes and 25 seconds".

But if you import that data to Excel or Google Sheets, you cannot do math on those numbers (like taking the average), because those apps will interpret those numbers as *4 minutes and 30.5 seconds", "4 hours and 45 minutes", and "4 hours and 25 minutes". (Well, actually, they're really being interpreted as timestamps, but the math would still mostly works as if it's durations in hours, minutes and seconds anyway.) I don't mention Numbers here, since Numbers doesn't seem to have the ability to do that kind of math on timestamps.

The right thing to do here would be something like this (which would work in Excel and Sheets):

4:30.5 (or 0:04:30.5)

0:04:45

0:04:25

(leading 0 added in "04" only for cosmetic purposes)

Ofc if it so happens that none of lap times has fractional seconds, it would work fine too:

4:30

4:45

4:25

Even though technically this is interpreted "4 hours 30 minutes" (or 4:30 AM), "4 h 45m", "4 h 25m", it doesn't matter because you can take the average of those durations/timestamps in Excel and get the right answer, regardless of whether Excel thinks the input/output are in hours and minutes or minutes and seconds.

I just don't think Garmin cares all that much tbh. That's why I don't use CSV exports for any kind of analysis. I prefer using runalyze.com (which syncs with Connect) to look at both individual activities and aggregate data, or using fitfilefileviewer.com to analyze the original data from activity FIT files.

DJGShef said:

scrolling down through the data to the end

But the infinite scroll implementation in the Connect website activity list sucks, as you can't jump to any point in the list instantly. The scrollbar / scrolling region only covers the handful of activities that are currently loaded / rendered. If you have a few years worth of activities, you will be scrolling for a long time.

(inb4 that's how infinite scroll works - yeah, for most sites/apps, but yet some sites/apps like strava, stryd and spotify figured out how to do this properly, so when they show users certain long lists, they have both infinite scrolling/lazy loading *and* they offer the ability to jump to any point in the list.)

DJGShef said:

the accurate time is under elapsed time or moving time

DJGShef said:

The 'Time' field is to the nearest minute

But ofc elapsed time and moving time aren't equivalent to the (total) time field....

Activity Type,Date,Favorite,Title,Distance,Calories,Time,Avg HR,Max HR,Aerobic TE,Avg Run Cadence,Max Run Cadence,Avg Pace,Best Pace,Total Ascent,Total Descent,Avg Stride Length,Avg Vertical Ratio,Avg Vertical Oscillation,Avg Ground Contact Time,Avg GAP,Normalized Power® (NP®),Training Stress Score®,Avg Power,Max Power,Steps,Body Battery Drain,Min Temp,Decompression,Best Lap Time,Number of Laps,Max Temp,Moving Time,Elapsed Time,Min Elevation,Max Elevation
Running,2025-10-16 16:21:53,false,"Glendale Running","7.07","610","1:22","118","125","3.2","169","190","11:34","8:14","177","174","0.82","10.0","8.2","274","11:35","194","0.0","192","295","13,740","-14","80.6","No","00:11:57","4","89.6","01:21:42","01:23:31","1,201","1,274"
Running,2025-10-13 16:18:32,false,"Mashpee Running","2.59","224","0:27","112","147","3.0","173","190","10:11","8:39","62","56","0.90","9.1","8.2","257","10:12","210","0.0","208","287","4,606","-7","60.8","No","00:06:00.4","2","77.0","00:26:12","00:30:08","94","130"
Running,2025-10-12 08:15:35,false,"Falmouth Running -Cape Cod 5K","3.16","278","0:31","132","146","3.6","181","192","9:42","8:43","39","36","0.91","8.6","7.9","248","9:42","222","0.0","219","262","5,542","-8","64.4","No","00:10:53","2","71.6","00:30:38","00:30:38","16","35"

The above is an export from the csv, saved through the Garmin Connect Web Activity export, with some recent running activity.  The times for the activities are highlighted in yellow.  The top one is actually a time of 1:21:50, the next one down is actually a time of 0:26;25, and the third is actually a time of 0:30:38.  It is obvious that these are now being rounded to the nearest minute, and this has only begun to happen recently.  Is it possible for me to be able to obtain the resolution that was being exported for years before this changed?  I have long term data in spreadsheets, and do calculations using that field.  It would be very helpful to have this fixed.

Someone else mentioned the same problem: https://forums.garmin.com/apps-software/mobile-apps-web/f/garmin-connect-web/424133/export-only-giving-run-times-to-nearest-minute 

My 2 cents:

- Garmin CSV exports suck in general (see the linked thread for an egregious example - lap CSV export)

- One workaround is to use runalyze.com, a free site that syncs with Connect. From there, you can export activity summary data to CSV without the problem that you're running into. (If you have more than a couple of years of data in Connect, you may need to do some extra steps to export/import all the data from Connect to Runalyze.)