Training & Behavior

A woman came to me last spring with a nine-month-old spaniel who could not be caught in the garden, hated his harness, and lost his mind when the doorbell rang. Three problems, three training plans, or so she assumed. We spent the first session teaching him to boop her palm with his nose. Forty minutes, a pot of chicken, one very silly-looking exercise.

Six weeks later the garden thing was gone, the harness went on without a fuss, and the doorbell had become a cue to run to her hand instead of at the door. She thought I had held something back. I hadn't. Hand targeting is simply the most quietly useful skill in reward based dog training, and almost nobody teaches it first.

Here is what it is. You present a flat open palm, the dog touches it with his nose, you mark and pay. But think about what he is doing physically: moving his whole body toward your hand, voluntarily, head up, weight forward. That single movement is the raw material for recall, for getting a dog off furniture, for walking a nervous dog past something scary, for repositioning at the vet. Not a trick. A portable steering wheel the dog operates himself.

Start indoors, somewhere boring. Hold your palm out about fifteen centimetres from your dog's nose, slightly to one side so he has to move a little. Most dogs sniff it within two seconds because hands smell like people and people smell like food. The instant his nose makes contact, mark with a word or a click, then deliver a treat from the other hand. Do not put food in the target hand. That is the single most common mistake and it teaches the dog to follow food rather than to offer the behaviour.

Reward schedule matters more than people expect. For the first eighty to a hundred repetitions, pay every single touch. Only once the dog is driving into your hand with obvious enthusiasm should you vary it, and even then not straight to one in five. Move to four out of five, then three out of five, jackpotting the best ones with three or four treats in a row. Variable schedules make behaviour durable, but only after the behaviour exists. Thin too early and you get a dog who tries twice and wanders off.

Sessions should be short to the point of feeling incomplete. Two to three minutes, ten to fifteen repetitions, then stop while the dog still wants more. Four two-minute sessions across a day beat one twenty-minute grind. Dogs consolidate learning in the gaps between sessions, not during them.

Raising criteria without breaking it

Once the nose touch is reliable in one spot, change one thing at a time. Only one. Hand lower so he dips his head, then higher. Present it sitting on the floor, then standing, then walking. Then take three steps back so he has to travel. That travelling version is where hand targeting becomes a recall you can actually use, because he is not vaguely coming toward you, he is coming to a specific point that ends in contact and payment.

Distance goes up in small jumps. Half a metre, then a metre, then two. Leap from two metres to ten and he stalls, and you have not found his limit, you have found your own impatience. Return to the last fluent distance, get three clean repetitions, then split the difference.

Distraction is the hardest variable and comes last. A dog who targets perfectly in the living room may ignore you in the garden at dusk with a blackbird in the hedge. That is context, not disobedience. Behaviour does not generalise the way we assume, so you re-teach it in the kitchen, the garden, the pavement, the park at 7am and the park at 5pm. Each location costs a few easy repetitions to reload. After five or six, most dogs work out that the palm means the same thing everywhere.

Regression is normal and it looks specific. A dog at 95 per cent suddenly drops to 60. He hesitates, offers a sit instead, sniffs the floor. Nine times out of ten one of three things has happened: you raised two criteria at once, the environment got harder without you noticing, or the dog is tired, sore or over-aroused from something earlier. The fix is boring and it works. Drop back two steps, pay every rep for a session or two, rebuild. Do not repeat the cue louder or stand there waiting him out, because repeating a cue the dog cannot currently answer teaches him it is optional.

Where it earns its keep

The garden-catching problem solves itself because the dog now has a specific, well-paid job that ends with him beside you rather than a vague sense that you want something. The harness gets easier because you ask for a touch, feed, briefly handle a shoulder, then touch again, building cooperative handling in three-second chunks instead of wrestling.

For the doorbell, targeting gave him something physically incompatible with charging the door. That is the real trick with any unwanted behaviour: give the dog a competing job that pays better than the thing he was going to do anyway. Shouting gives him nothing to do instead. A trained nose touch does.

It shines with worried dogs too. A shy dog who will not walk past the recycling bins can often be moved past them one touch at a time, because targeting keeps his head up and gives him a rhythm to follow. You are not dragging him. He chooses each step. That distinction matters enormously to a fearful animal, and it is why I lean on this rather than anything involving pressure, corrections or the tired old dominance framing. A prong collar or a leash pop teaches a dog nothing about what to do. It only tells him he was wrong, which he already suspected.

Two honest caveats. Some dogs, particularly those grabbed or smacked in the past, find an approaching palm genuinely alarming. If yours ducks, freezes or backs away, do not push it; teach a target stick or a sticky note on the wall instead. And if you are dealing with real aggression, severe separation distress, or a dog who has bitten, hand targeting is a useful tool inside a larger plan but it is not the plan. Get a credentialled behaviourist involved alongside your vet, because pain sits underneath a surprising number of behaviour cases.

What I like most is how forgiving this exercise is. No timing is so bad, no session so clumsy, that it does harm. It becomes the place where nervous owners learn to mark and pay and read their dog, and by the time they are fluent they have accidentally built a skill they will use every day for a decade. Start tonight. Two minutes. Flat palm.