I love the idea that a marathon doesn't finish at the line. Not really. The finish is where you get the receipt.

Alex Yee ran his first London Marathon and posted a ridiculous debut time of 2:11:08. But even at that level, the last miles still did what the last miles always do. They asked questions. Later, he compared his COROS data with Eliud Kipchoge. Kipchoge's response was perfect because it was both reassuring and quietly savage in that way only the best can manage: it's normal, it's your debut, and next time you learn from it.

That's the point of data when it's used well. It's not there to make you feel guilty about slowing down, or to turn every run into an exam. It's there to tell you what happened, where it happened, and what you can do about it.

Kipchoge put it in a way I keep coming back to. Data tells a story of how you like to run. Where you go fast, where you go slow, where you hold it together, and where it starts to slip. If you struggle at the end, the data can show you what you're struggling with. Which is basically the most useful thing anyone can say about a marathon.

So, what should you actually look at after race day if the goal is to run a better one next time?

1) Effort over time: use your zone distribution

The cleanest way to see whether you paced the marathon well is to look at how your intensity shifted across the race. Most watches will show this through heart rate zones, and often pace zones too. You're not hunting perfection here. You're hunting a pattern.

A well-paced marathon usually looks like controlled aerobic work for most of the race, with a gradual drift towards harder effort late on. That drift is normal. What you don't want is the early spike. If your zone distribution shows you spending too long too early in the higher zones, you didn't "go for it". You borrowed from the end of the race without checking the interest rate.

The useful question is simple: did your effort stay steady while you moved through the miles, or did it jump around? If it jumped, the next step is figuring out why. Terrain, wind, crowd adrenaline, gels missed, going off plan, all of it leaves fingerprints.

2) Learn from every split, but don't get lost in them

Splits are where the marathon starts to talk back. Not because one slow mile ruins you, but because of what the trend is doing.

Zoom in to each mile or every 5K and look for the moment your pace began to fade. Then cross-check it with heart rate. This is the important bit.

• If your pace dropped but your heart rate stayed maintained, you're looking at fatigue showing up in the legs. You're still trying, but the output isn't there anymore.
• If your pace dropped and your heart rate also dropped, you might be seeing a proper bonk moment, or a moment where you stopped being able to drive the effort.

The goal isn't to beat yourself up. It's to find the tipping point. The mile where it started to cost you, not the mile where it finally fell apart. That tipping point is gold for training, because it tells you what you need to be ready for next time.

3) Form under pressure: cadence and stride length

This is where marathon data gets genuinely interesting, because it moves beyond "you slowed down" and into "how you slowed down."

With COROS, you can view cadence and stride length breakdowns in your activity summary. If you spot a steady drop, you've basically got a warning light from your own mechanics.

A few simple reads:

• A gradual drop in stride length can be muscular fatigue. You're still turning over, but you're losing power and extension.
• A decline in cadence can be a sign you're overstriding, tightening up, or struggling to keep turnover when things get hard.

Neither is "bad" on its own. They're just signals. And the reason they matter is because they tell you what kind of work you might need more of. If stride length collapses late, that can cue strength and durability. If cadence drops, that can cue efficiency, rhythm, and staying relaxed when you're tired.

It also gives you something practical to use mid-race next time. Because when you're deep in the final 10K and your brain is offering you the worst suggestions imaginable, cadence is often the simplest lever to pull. Not a huge change. Just enough to keep the pattern alive.

The runner's way to use all of this

Here's the trap with data. You can stare at it for hours and still not change anything. The win is picking one or two takeaways that actually affect training.

Look at your race and decide which story it's telling:

• Did you go too hard early? Then next cycle is about restraint and aerobic strength.
• Did you fade late with effort still high? Then next cycle is about durability, long-run structure, and strength.
• Did your mechanics fall off a cliff? Then you build a body that holds form when tired.

That's how you turn a marathon into a better marathon. Not by obsessing over every chart, but by listening to the bit of the story you're currently trying to rewrite.