Meet Baelor.
He's the reason FetchCoach exists. This is his actual training journey — real check-ins, real notes, honest progress. Nothing curated.
Training in progress. Check back soon!
I got Baelor in late January 2026. I started building FetchCoach within a week.
I was already using other dog training apps when I brought him home. They weren't cutting it — not because they were bad, but because they were fundamentally the wrong thing. Static how-tos. Step-by-step guides written for a generic dog in a generic situation. That's not training. That's content.
What I actually needed was something dynamic. A trainer that knew my dog — not dogs in general, but Baelor specifically. What we'd been working on. What had clicked and what hadn't. What I'd asked about last Tuesday at the dog park. Something that built a real picture of him over time and used it to give better guidance each session.
That tool didn't exist. So I built it.
FetchCoach is the app I went looking for and couldn't find. Every feature in it came from a real problem I ran into training Baelor — the skills system, the check-in format, the voice coaching. Baelor is the first user. He's also the ongoing case study. What you're looking at below is my actual account, my actual training data. Nothing staged.
Baelor's saved skills
Every skill I've added to Baelor's training plan, how many sessions we've logged, and when we started and last checked in. Pulled live from my account.
The actual sessions. Wins, struggles, and notes.
Dog training is not linear. Here's what real progress looks like from the skills we've logged the most sessions on. These are my real notes. Nothing edited for comfort.
Name response
Marker Word
Sit
Where we actually test it.
Baelor's recall got built indoors, in our backyard, and in low-distraction parks. But recall that only works in controlled conditions isn't recall. The Elbow River trail near Bragg Creek is where we find out if it's real.
The distractions there are stacked: other dogs, cyclists, kids, water, wildlife, and a hundred interesting smells competing for his attention every ten metres. Off-leash trail work is the hardest context to proof a skill in — and the one that matters most if you want to actually let your dog run free.
Here's what the process looks like for recall at distance:
- Build value on a long line first. Baelor drags a 15-metre line while I call him randomly, reward heavily, and release him back to sniffing. The line is safety — it means the recall always succeeds.
- Proof around specific distractions. We add one distraction at a time: moving water first, then bikes, then other dogs at distance. Don't try to proof everything at once.
- Off-leash starts close, not far. First real off-leash reps happen within 3–5 metres. I build distance only after he's demonstrated the pattern at short range under distraction.
We're not all the way there yet. "Come" works reliably on trail when there's no other dog in view. Add another dog — especially a running one — and the success rate drops. That's what we're still working on.
See the recall training guide →
Session photos coming soon
Session photos coming soon
Session photos coming soon
Not everything is solved.
Baelor is 3 months old. I'm not going to pretend he's a finished dog.
Here's where we actually are as of May 2026:
Recall around other dogs
Name response at home: solid. Recall on a quiet trail: reliable. Recall when another dog is running nearby: not there yet. Off-leash recall around active distractions is the hardest skill to proof, and we haven't proofed it. We're still on a long line in those contexts.
Crate training
It's on the skill list but we haven't started logging sessions. He tolerates the crate — he doesn't love it yet. Getting him to the point of choosing to go in on his own is a few weeks of deliberate work away.
Name recognition while overstimulated
His own notes from May 3: "still need to work on name recognition while overstimulated ie at the dog park." Dog parks are sensory overload and the hardest context for attention work. We haven't solved this.
FetchCoach doesn't work because Baelor is an exceptional dog — he's a normal pup with normal pup problems. It works because the skill-gap frame helps me understand why he's failing in these contexts, and the coach gives me a concrete protocol to close each gap. The progress is real. So are the gaps.
Training Baelor shaped every product decision.
Why skills exist as a separate system — not just chat history
Early versions of FetchCoach were just chat. You asked, the coach answered. The problem: every session started from scratch. The skills system exists because I kept wishing there was a "remember we're working on recall and here's how it went last time" layer. Now there is.
Why voice coaching matters more than I originally thought
I added voice because it seemed like a useful feature. I kept it because it became the only way I actually used the app during training sessions. When both hands are occupied and your attention is split, pulling out a phone to type isn't an option. Talking is.
Why honest check-in categories beat arbitrary rating scales
The three check-in options — Win, Still Learning, Got Feedback — came from real sessions with Baelor. What actually captured how a session went was "this worked", "we're making progress but not there yet", or "something happened I want the coach to know about." The third one forces you to write down what went wrong — that note becomes context for the next session.
Week 3: taking it out of the living room.
Baelor has graduated Week 2. The Week 3 sessions take each Tier 2 skill out of training conditions and into real environments — chained with a foundational Week 1 skill. This is where the training becomes the behavior.
Your dog deserves the same.
$5/mo locked forever for the first 200 founding members. I built this for Baelor. Use it for yours.
Claim Your Founding Spot — $5/mo3 messages free · No account needed · Cancel anytime
📅 Full Training Timeline · This Week's Sessions · About Jason & FetchCoach · Browse All Skills · Pricing