What 201 Facebook reels taught me about short hiking videos

I make long films about hiking alone, mostly for YouTube, and have for about ten years. This year I started turning them into short vertical videos for every platform. On October 9 I had Claude, the AI I work with, pull the numbers on every reel I've ever posted there and tell me what works.

Why Facebook, in my own words

Facebook wasn't really that important to me before this year. I completely disregarded it for years, and I kind of regret that, because it's a really powerful place to post, and I think the community there is actually pretty good. It goes along with this idea of making one piece of content and then sharing it in lots of different places. I did all of the hard work but just didn't do the easy step of distributing it to all these places, like Facebook. And Facebook is already making me money and growing in a way that YouTube and Instagram really aren't. So it's interesting to see. I really like Facebook, and posting reels there just makes total sense.

Why look at the numbers now

Two reasons:

• Facebook is a discovery channel for me. 92 to 94% of the watch time on my reels comes from Facebook recommending them to strangers, not from people who follow the Page. So each reel lives or dies on its first seconds with people who've never heard of me.
• A Facebook creator challenge rewards posting reels often. Before posting a lot more, I wanted to know which reels to post.

How the analysis worked, in 6 steps

1. Export. Claude downloaded three read-only reports from Meta Business Suite: one row per post, one row per video (watch time and where it came from), and the Insights content table (saves, follows).
2. One table. It matched them into one row per reel, 201 in all, each with its link. My Page has nothing before November 2025, so that's the whole history.
3. Label. Each short got a format (story moment, trip montage, place clip, map reel and so on), a length, whether the caption says "alone" or "solo", and how many shorts went up that day.
4. Compare. For each group, the median views: the reel in the middle, so one viral hit can't skew it.
5. Rank. The findings went into a ranked list of every short I have.
6. Post from the top. That list now decides what goes up next.

All views below are Facebook-only lifetime views as of October 9. The newest reels have had days to collect views, the oldest about eight months. Claude did the pulling and the counting; KraigOS, the system Claude and I built to run my work, keeps the clip library the results feed into.

One real moment beats a montage

My most-viewed short isn't a trip summary. It's 58 seconds of one night alone in the Three Fingers fire lookout: 435K views, 2,854 saves and 399 shares, and people watched 20 seconds of it on average. It's cut from my film about that trip, and the route is in my Three Fingers guide.

Reels built on one moment or scene had a median of 28.9K views, the best of any big group, and people watched 36% of each one on average. A rainstorm on the Ak-Suu Traverse in Kyrgyzstan got 104K, with people staying almost 18 of its 47 seconds. (My Ak-Suu guide.)

Clips titled with just a place, like "Hiking in the Azores" or "Hiking in Nepal", did worst: a median of 10.7K. The exception made a claim. Best Hike on the Island of Madeira got 185K. "Best" gives a stranger a reason to watch that "Hiking in Madeira" doesn't. (My Madeira guide.)

The split-map route reel is the strongest new format

Split-map reels show my route drawing across a map on one half of the screen while my footage plays on the other. Only two have gone up on Facebook, so this is early, but both did well:

• Hornstrandir, Iceland: 158K Facebook views in its first ten days and 1,426 saves across Facebook and Instagram, the second most saves of any short after the lookout. (The guide, the film.)
• K2 Base Camp, with my AI voice narrating: 39K after two days. (The guide.)

For comparison, a 15-second montage of the Alta Via 1 made from the same kind of footage got 31K, and people watched 2.6 seconds of it.

"Alone" in the caption

The 17 shorts whose caption says "alone" or "solo" had a median of 20K views. The other 69 had 12.6K. Seven of my top ten shorts say it, including Hiking the Italian Alps alone (399K), 80 miles alone on the Skye Trail (171K) and 6 days hiking 85 miles solo in the Lofoten Islands (104K). It's true of almost every trip I film, so now it goes in wherever it fits. (Guides: Alta Via 1, Skye Trail, Lofoten.)

Posting in bulk hurt, and repeating a topic hurt more

On February 9 and 10 I put up 43 shorts in two days. Shorts posted on days with five or more had a median of 9.5K views. Shorts posted on days with one to four had 19.6K, about twice as many.

Same-topic reels split the audience the most. Three Azores reels went up within three minutes: the first got 119K, the second 8.4K and the third 7.1K. A Salkantay Trek montage posted the next morning, a minute after another Peru reel, got 2.2K, my lowest short. Nine of my ten weakest shorts came from those two days.

Even inside the dump, a strong reel still broke through: the lookout and Madeira reels both went up on February 10. A strong moment survives bad timing. A generic one doesn't. (Azores guide, Salkantay Trek guide.)

30 to 45 seconds, and seconds watched matter most

Shorts of 31 to 45 seconds had the best median, 16.9K. Clips of 15 seconds or less and cuts of two or three minutes did worse unless the story carried them; the 178-second Skye Trail recap is the exception, at 171K.

Views went up with the average number of seconds people watched more than with the percent of the reel they finished. So a short that's too short to hold anyone for long has a ceiling, even if most people finish it. The 3-second view rate told us nothing.

Long films posted as reels drew the most views

Facebook labels every video on a Page a reel, including full-length films. In February I uploaded my YouTube back catalog there. Those 115 long films drew 4.19 million views in total, more than all 86 shorts together (3.23 million).

It's top-heavy. My 17-minute Alta Via 1 film has 1.04 million views on its own, and the 14-minute K2 film has 778K. The typical long film did less than the typical short (8.8K against 12.9K). The hut-to-hut and expedition films are what Facebook loves most.

Sponsored reels reach people but don't hold them

A sponsored reel from my Dents du Midi trip got 147K views, but people watched 2.9 seconds of it on average, and only 2 shared it. Another sponsored travel reel got 57K views at 3.3 seconds. Across all 18 sponsored shorts, about 0.05 people shared per 1,000 views. For everything else it was 0.45, about ten times as many.

A tip for other creators: the Business Suite "Views" column

For a reel crossposted from Instagram, the Views column in Meta Business Suite adds the Instagram views in. My Hornstrandir reel shows 270,956 views there but 157,995 on Facebook alone. Saves, follows and link clicks for crossposts look combined too. If you're judging Facebook, use the post export's Facebook-only number.

Follows aren't a Facebook signal

Facebook reported 0 follows on every one of the 58 native reels that had a follows number. Follows only showed up on Instagram crossposts, which suggests they're Instagram's. So we don't judge a Facebook reel by follows.

What we changed

What surprised me is how large the hiking community is there. There are a lot of people looking for advice on where to travel and how to hike, so the community is pretty good.

I'm doing research on formats and trying to learn how people make photo posts, carousels, reels and long form, and what that looks like on Facebook, because I'm used to the other platforms. So I'm watching a little on Facebook just to understand what works there, how people share and what people expect. The short vertical videos, the reels, do really well on Facebook. I think that's natural, because people are probably watching on their phones. And like I said, I'm making a ton of vertical shorts that can be posted in lots of different places. So why not?

Here's what changed after the analysis:

1. Every short is ranked. Claude scored every clip in my library, 67 of them, best to worst on how we expect it to do, using these findings. The list is re-ranked as new numbers come in.
2. We post from the top. The best-ranked shorts go up first.
3. One or two a day, never a dump. No route repeats within about three days, and no caption repeats wording the Page just used.
4. Map-route reels and one-moment reels come first. Place-name montages go last.
5. Weak ones get re-cut. Most map reels run 55 to 71 seconds, so their next version is a 30 to 45 second cut.
6. "Alone" goes in the caption when it's true.

What's next

More map reels for the routes in my guides, the 30 to 45 second cuts, and a re-rank each week as the numbers come in. In a few weeks there'll be enough new reels to check whether these patterns hold.

FAQ

How often should I post reels on Facebook?

On my Page, reels posted on days with one to four shorts got about twice the median views of reels posted on days with five or more. Posting several reels on the same topic within minutes hurt the most. I now post one or two a day.

What's the best length for a Facebook reel?

For my hiking shorts, 31 to 45 seconds had the best median views. What matters more is how many seconds people actually watch, so a reel needs a reason to keep watching, not just a short length.

Does the Views number in Meta Business Suite include Instagram?

For reels crossposted from Instagram, yes: the Views column in Business Suite adds the Instagram views. Use the post export's Facebook-only views to judge Facebook.

Should I post my long YouTube videos as Facebook reels?

They drew more views in total than all my shorts, but a few big films carry that total, and the typical long film got fewer views than the typical short. Films about long hut-to-hut and expedition trips did best.

Do sponsored reels do worse on Facebook?

Mine reached plenty of people but held them for about 3 seconds and were rarely shared: about a tenth of the share rate of my other shorts.

How did you analyze your Facebook reels?

Claude downloaded read-only exports from Meta Business Suite, matched them into one table with a row per reel, labelled each short by format, length, caption and posting day, and compared median views for each group.