Apple Fitness called my first run Hard. I still didn't know what to do next

My first 1.67 kilometres

On July 2, 2026, I went out for my first run. I started at 8:50 in the morning and stopped 11 minutes and 54 seconds later. I had covered 1.67 kilometres.

I stepped off the track, took out my phone, and opened Apple Fitness while I was still sweating like crazy. The screen gave me plenty to look at: an average pace of 7'07"/km, an average heart rate of 160 bpm, 152 steps per minute, 219 watts, and 130 active calories.

The heart-rate number made sense in a very physical way. It felt as if my heart was trying to jump out of my chest. The cadence number meant absolutely nothing to me.

Then I saw the effort rating: 7, Hard. I remember thinking, half exhausted and half amused, “Is there no level above Hard?”

A mid-July run exposed the real problem

By July 15, I had gone out again after quite a bit of psychological preparation. I had also asked AI to help me read the data from my first run, and cadence was the number I decided to watch.

For the first part of the run, I could still pay attention to my steps. As the distance increased, that attention disappeared. I was tired, sweating heavily again, and using most of my concentration just to keep moving. Asking myself to monitor and correct my cadence at the same time wasn't realistic.

That run was 2.44 kilometres at 7'40"/km. My average heart rate and cadence were both 159. Most of the run still felt hard, although a short stretch felt noticeably easier. I was drenched afterward; during the slow walk, I felt surprisingly relaxed.

The dashboard I had to take apart

Years ago, I took over the first version of an admin dashboard for setting up campaigns. It had plenty of features, with entry points scattered across the product. Users needed a thick manual and complicated tutorial videos just to create one campaign.

My proposal was to remove those entry points and rebuild the workflow as a timeline. It wasn't an easy judgment to prove in advance, so I had to use the questions coming from users and the available data to convince the team that fewer choices could work better. After the new flow went live, consultation volume dropped sharply. Visits to the manual and tutorial videos dropped too.

The one metric I could act on

I stopped trying to understand everything on the Fitness screen. Cadence was the one number I had actually tried to change during the July 15 run, so I focused there.

The rule I set for myself: hold a stable pace range, gradually increase cadence until it reaches a reasonable level for that pace, and only then think about running faster. Changing one thing at a time felt safer and, more importantly, possible to follow when I was tired.

What I built into Health Run

That rule became the gradual guidance in Health Run. When my pace has been stable, the watch might say: “Keep your current pace and increase your cadence slightly.” The metronome gives me the beat, and periodic voice reminders bring my attention back when it starts to drift.

I am basically following a rule I wrote for myself. That sounds slightly ridiculous, but it works much better when the watch remembers the rule for me. Afterward, the cadence profile lets me see whether I actually held the rhythm across several runs.

What I want to improve now

The annoying thing about building a running app is that I can't test it by sitting at a desk and clicking around. My legs are part of the test suite. From July through early August, debugging meant going outside two or three times a week and running under the summer sun.

Between August 9 and August 22, I changed the plan to a series of one-kilometre adaptation runs. They were continuous runs, without walking, at roughly 8'40"–9'20"/km. I covered 11.6 kilometres in total. My heart rate was usually around 130–135 bpm, cadence around 158–160 spm, and the easy-running share was often 50–60 percent.

The lower heart rate wasn't magic. I had slowed down and stopped each run at about one kilometre. If I ran beyond 2.5 kilometres now, I expect it would climb again, although probably not straight back to 160.

I finish these short runs, check the easy-running share, and see whether the rhythm held.