Three weeks ago your puppy sat every single time you asked. Now you say it in the kitchen and they look straight past you at the bin. You've tried saying it louder, holding the treat higher, and repeating the word until it stops meaning anything. Somewhere in there, someone told you they're testing you.
They're not. A puppy not responding to training is showing you a mechanical fault in how the training was built, and there are only four faults it can be. Luke Buchanan, Owner of The Toe Beans Co and Sydney's Puppy Trainer, works through the same four with nearly every owner who books a session convinced their dog turned stubborn overnight. None of the four is defiance, and the fix is almost never "be firmer".
Why your puppy is not responding to training
A puppy not responding to training has hit one of four things: they're over threshold and can't take information in, the criteria moved too fast, the reward stopped meaning anything, or the cue was never actually learned in that context. That's the whole list.
The reason owners reach for stubbornness instead is that the dog training industry is roughly twenty years behind human behavioural science, and "he knows what he did" is the clearest example of the gap. No teacher would watch a seven-year-old do long division on Monday, fail it on Friday in a noisy assembly hall, and conclude the child was being defiant. They'd ask what changed in the room. Dog training hasn't fully made that move, which is why plenty of what gets repeated as common sense is quietly wrong about what puppies can and can't do.
Cause one: your puppy is over threshold
A puppy above roughly a five out of ten on arousal physically cannot take in new information, so nothing taught at that level sticks. Threshold is the point where a dog's arousal makes calm behaviour, and therefore learning, impossible.
Picture your puppy's arousal on a one-to-ten scale. At one or two they're settled and genuinely receptive. Three to five is what most owners call calm in an excitable dog, and it already isn't calm enough for much to land. From about seven upward the puppy hasn't chosen to ignore you, they've gone offline. In my sessions the most common thing I see is an owner describing their puppy as a three or four when the dog in front of me is sitting at a six or seven, and that gap is where most home training quietly dies.
Bring the number down before you ask for anything. A calm, silent hold under the collar until they settle, a quiet timeout in a boring room, or simply waiting it out. No talking, no repeating the cue, no lead pressure. Then start the session.
Cause two: the criteria moved too fast
Distance, duration and distraction are three separate skills, and adding more than one at a time is what breaks a cue that was working yesterday.
Here's the version I see constantly. The puppy learns to sit in the kitchen with you standing right there. It's solid, so on Saturday you ask for a sit at the park from two metres away with a Kelpie sniffing about ten metres off. That's distance, distraction and a new environment introduced at once, and the sit falls apart.
Add one variable at a time and only move up once the current level is genuinely reliable. If it breaks, drop back to the version that worked and finish there, because ending on a success is worth more than squeezing out one more rep. That's also why the core commands are worth building slowly rather than rushing to prove them in public.
Cause three: the reward stopped meaning anything
A reward only works while the puppy actually wants it, so a reinforcer repeated forty times, or one the dog gets free from a bowl anyway, has stopped being a reinforcer.
Kibble the puppy grazes on all day is not payment, and neither is the twentieth identical treat in a ten-minute session. There's a subtler version that catches owners who've done everything else right: a dog that always gets a treat is coming back for the treat, not for you. Sometimes food, sometimes a scratch, sometimes a toy, sometimes nothing special at all. That variability is what builds a response that holds up when you've got no food on you.
Check one related fault. If you taught the sit with a treat lure over the nose and never faded the hand, the cue is your hand, not the word. Empty your fingers and see whether the puppy still responds. Plenty don't, and that's not disobedience, it's an unfinished piece of reinforcement work.
Cause four: the cue was never learned in that context
Dogs don't generalise the way people assume, so a cue learned in one room is often not the same cue anywhere else.
To a sixteen-week-old puppy, "sit" isn't an abstract word. It's closer to "sit, in the kitchen, on lino, facing the counter, with you standing upright in front of me". Change the surface or the room, or sit down on the floor yourself, and you've handed them a new problem. Generalising a cue means rehearsing it deliberately in the hallway, the garden, the lift and the footpath outside, one setting at a time.
Recall is where this bites hardest, because it's the cue you need at the moment the world is most interesting. Call once, and if they don't come, walk away and wait rather than repeating it. Never call when you don't believe they'll come. Never call your puppy and then do the thing they hate, go and collect them calmly instead. A recall must never end the fun, which is the mistake that quietly ruins more recalls than anything else.
How to tell which one you're looking at
| What you're seeing | Most likely cause | First move |
|---|---|---|
| Works at home, falls apart outside | Over threshold, or context | Read the arousal level before you judge the cue |
| Worked yesterday, gone today | Criteria moved too fast | Drop back a level and rebuild |
| Responds slowly, or only sometimes | Reward has gone flat | Change the reinforcer and randomise it |
| Only responds with food visible | Lure never faded | Empty your hand and retrain the word |
The diagnosis matters more than the technique, because the same behaviour has four different fixes depending on what sits underneath it. That's the core problem with training advice written for no dog in particular.
Mistakes that make it worse
Repeating the cue is the big one. Saying "sit, sit, sit, SIT" teaches your puppy the word requires nothing until the fourth time, so you've trained a delay into the response on purpose.
Getting louder or adding physical pressure does two things, neither useful. It pushes the arousal number up, which is the opposite of what a puppy already over threshold needs, and at The Toe Beans Co it's off the table. We never use pain, fear, force or aggression. Aversive Correction (+P) is not part of the method in any form. Access Withdrawal (-P), calmly removing what the dog wants, does the job without the fallout.
Training while you're frustrated is the one owners find hardest to hear. Puppies read tension in your body long before they read the word, and a tense handler reliably produces a slower response. The most common mistake I see is pushing through one more repetition to end on a win, and getting a worse rep, a flatter dog and a slightly damaged cue instead.
When to get someone in
Work through the four causes honestly for two to three weeks. If you've dropped the criteria, changed the reward, trained below threshold and rehearsed the cue in new places and still get nothing, something else is going on.
Get help sooner if there's growling, snapping or freezing involved, if your puppy is shutting down rather than switching off, or if things are getting worse. Those aren't training-mechanics problems and they don't respond to more repetitions. For dogs of thirty-two weeks and older, one-to-one behaviour training in your own home is usually the fastest route, because the trainer sees the actual room, the actual routine and the actual handler. For younger puppies, puppy school covers the same ground in a group.
Want help putting this into practice?
The Toe Beans Co runs a free SKOOL community where Sydney dog owners get access to training guides, Q&As, and direct support from Luke. It's free to join.
Upcoming puppy schools in Sydney
Luke runs regular puppy schools across Sydney, with venues in Paddington, Edgecliff, Zetland, Alexandria, Marrickville and Neutral Bay. All six are taking bookings, and anyone in Sydney can attend any location.
Check upcoming dates and book your spot →
Frequently asked questions
Q: Is my puppy being stubborn or just not listening?
A: Neither, in almost every case. A puppy that stops responding is over threshold, has been moved through the criteria too quickly, has stopped valuing the reward, or never learned the cue in that particular setting. Stubbornness isn't a mechanism you can train against. All four of those are.
Q: How long does it take to fix a puppy not responding to training?
A: Most owners see the response come back within one to two weeks of correcting the actual cause. The rebuild is quick because your puppy isn't unlearning anything, you're retraining at a level they can succeed at and building up one variable at a time.
Q: Where can I get help with a puppy that's stopped responding in Sydney?
A: The Toe Beans Co runs puppy school across six Sydney venues, including Paddington, Edgecliff and Marrickville, plus one-to-one in-home training for dogs from thirty-two weeks. In-home work covers roughly a forty-five minute drive of the Harbour Bridge.
