How to Train Your Dog: A Complete Beginner Guide

Most training advice tells you how to train — reward the right thing, get your timing right, keep it short. That advice is correct, and we have written it up as eight dog training tips.

This is the other half: what to teach, in what order, and how long each piece actually takes.

Because the usual failure is not bad technique. It is teaching "sit" in the kitchen, deciding it is done, and then discovering the dog has no idea what the word means in a park.

Before any command: two foundations

Skip these and everything after them is harder.

Their name means "look at me". Not "come here", not "stop that" — just look. Say the name once, and the instant their head turns, mark and reward. If you have been using their name to interrupt bad behaviour, it has become a mild warning sound, and you will need to rebuild it.

A marker word means "that exact thing earned a treat". Pick one sound — "yes", or a clicker — and use it only for this. The marker solves a timing problem: you cannot get a treat into a dog's mouth in the half-second that matters, but you can make a sound. The sound bridges the gap.

Spend two or three days on these before teaching anything else. It feels like a detour and it is not.

The six commands worth teaching first

In this order. Each one leans on the one before it.

1. Sit. The easiest win, and the one that teaches your dog that offering a behaviour makes good things happen. Hold a treat at their nose, move it slowly back over their head; as the nose goes up the rear goes down. Mark the moment it lands.

2. Down. Harder than sit, because lying down is a more vulnerable position. From a sit, lure straight down to the floor between their front paws. If they pop back up, you moved too fast.

3. Come. The one that matters most and gets ruined most often. The rule: never call your dog to something they will dislike. If you call them and then clip the lead on to end the fun, or call them for a bath, the word stops working. Go and get them for those things instead.

4. Stay. Build it on three separate axes, one at a time — duration, distance, distraction. Two seconds at arm's length in your kitchen, then five, then ten. Only when duration is solid do you add a step backwards. Adding all three at once is why stay fails.

5. Leave it. The safety command. Closed fist with a treat inside, let them sniff and paw, mark the moment they back off, then reward from the other hand. The lesson is that ignoring a thing produces something better.

6. Loose-lead walking. The hardest, because pulling has been rewarded by forward motion every single walk. The mechanic is simple and requires real patience: when the lead goes tight, stop. Not a correction — just stop. Move again when it slackens.

The step almost everyone skips

A command your dog performs in your kitchen is not trained. It is a trick that works in a kitchen.

Trainers call the missing step generalisation, and dogs are unusually bad at it. A dog who sits reliably at home genuinely may not know what "sit" means in the garden, because to them the cue includes the room, your posture, and the treat bag on the counter.

So for every command, work through:

• A new room, then the garden, then the street
• With distractions — another person, a passing dog, food on the floor
• Without a visible treat, rewarding from a pocket after the fact
• From different positions — sitting down, with your back turned, from another room

This is where the weeks go, and it is the difference between a dog who knows six words and a dog who is genuinely trained.

Consistency is the whole game here, which is exactly what daily tracking is for — DearPup keeps the streak visible so training does not quietly stop after the first enthusiastic fortnight.

What the evidence says about method

This part is not a matter of opinion any more.

The American Veterinary Society of Animal Behavior's 2021 position statement recommends reward-based methods for all dog training, and states that aversive methods — electronic collars, prong collars, choke chains, leash corrections, physical or psychological punishment — should not be used under any circumstances.

The reasoning is that there is no evidence aversive methods work better, and several studies find reward-based training more effective. A 2020 study on training method and welfare found dogs trained with aversive methods showed more stress behaviours during training — lip licking, yawning, lowered posture, raised paw — and the AVMA's write-up of the veterinary behaviourists' position is blunt that these tools damage both welfare and the bond with the owner.

If a trainer proposes a shock or prong collar, that is a reason to find a different trainer.

Puppies: start earlier than feels safe

The single most consequential window closes early. The critical socialisation period runs from roughly 3 to 14 weeks, and AVSAB recommends puppy classes from 7 to 8 weeks — deliberately before the vaccination course finishes.

That surprises people, so here is the reasoning: more deaths in dogs under three are attributed to behavioural problems than to infectious disease. Waiting until 16 weeks to socialise trades a small, manageable infection risk for a large, permanent behaviour risk.

Sensible precautions during that window, per AVMA guidance on socialisation: classes that require proof of vaccination, homes of known healthy vaccinated dogs, carrying your puppy through areas where unknown dogs may have toileted, and making vet visits a positive experience.

Feeding matters alongside it — a puppy training twice a day is burning energy, and how much to feed a puppy is worth getting right rather than guessing.

A realistic timeline

Treat that as a shape, not a schedule. Breed, age, history and how many minutes a day you actually put in all move it. A rescue with an unknown past may spend the first fortnight just learning that you are safe and predictable, and that time is not wasted.

If your dog seems unable to concentrate anywhere, that is worth separating from stubbornness — dog anxiety presents as distraction and non-compliance far more often than owners expect, and no amount of repetition fixes it.

Common ways it goes wrong

Repeating the cue. "Sit. Sit. Sit-sit-sit." You have now taught that the real cue is the fourth one. Say it once, wait, and if nothing happens make the exercise easier rather than louder.

Training a tired or overstimulated dog. Ten minutes after a long walk is worse than ten minutes before one.

Fading treats too fast. Rewards should become unpredictable rather than absent. Intermittent reinforcement produces more durable behaviour than constant reward — the same mechanism that makes slot machines compelling.

Practising only in easy conditions. Covered above, and it is the big one.

Expecting the dog to work out what you want. Dogs do not reason about intentions. They connect what they did with what happened next, within about a second. If the connection is not obvious in that window, nothing is being taught.

The takeaway

Teach the name and the marker first. Then sit, down, come, stay, leave it and loose-lead walking, in that order. Then spend most of your time doing those same six things in harder places, which is the part that actually produces a trained dog.

Use rewards, not corrections — the veterinary consensus on that is settled. Start puppies before the vaccination course finishes. Keep the sessions short and frequent, and end them while your dog still wants more.