These three behaviors — jumping, mouthing, and pulling — are probably in the top five reasons dogs end up surrendered to shelters or labeled "bad dogs." And they have more in common than it seems.
All three are high-energy, social, attention-driven behaviors. All three are often accidentally trained through the owner's response. And all three are fixed through the same underlying approach: remove the reward for the unwanted behavior, consistently reward an incompatible alternative.
The Common Thread: Attention Is the Reward
Jump on person → person reacts (pushes away, says "off," makes eye contact, even negative attention is attention) → behavior is reinforced.
Mouth hands during play → person says "ow!" and keeps playing → mouthing gets a response, play continues.
Pull on leash → dog reaches interesting smell or another dog faster → forward progress reinforces pulling.
In each case, the dog's behavior produces a consequence that keeps the behavior alive. The fix isn't to add punishment — it's to remove the reinforcing consequence.
Fixing the Jump
The dog jumps to get your attention and greeting. Give that attention and greeting for four paws on the floor, not for the jump.
When the dog jumps: turn away, cross your arms, zero eye contact, zero voice. When all four feet hit the floor: immediate, warm greeting. The dog quickly learns that the jump produces nothing, and four on the floor produces the greeting they want.
The failure mode is inconsistency — even one warm greeting of a jump resets the learning. Everyone the dog interacts with needs to run this protocol. See stop dog jumping on people for the full protocol.
Fixing the Mouthing
Mouth on skin: play ends immediately. Stand up, turn away, fold your arms, leave if necessary. Wait 10-30 seconds. Return to play. If mouth on skin again: repeat.
The dog learns that teeth on skin ends the game. The game is the reward. No teeth = game continues.
Two failure modes: (1) continuing to play through the mouthing because "it doesn't hurt that much," which tells the dog mouthing is acceptable; (2) yelling at the dog, which can be stimulating enough to look like a game response.
See stop dog mouthing hands during play for the detailed protocol.
Fixing the Pull
Pull on leash → stop walking completely. No forward progress when leash is tight. Wait until leash is loose (dog orients back, slack in leash), then walk forward.
The dog learns that pulling stops forward progress, loose leash allows it. This takes enormous consistency and patience — especially in the first weeks — but it works.
The failure mode: moving forward even slightly when the leash is tight, which tells the dog that enough sustained pulling eventually works. (See also: basic obedience training for dogs.)
The Consistency Problem
All three of these fixes require consistent application from everyone in the household and from strangers the dog interacts with. A strategy that works 80% of the time teaches the dog that the behavior works 20% of the time — and intermittent reinforcement is extremely durable.
If you can't control how everyone responds to the dog (guests who let the dog jump, strangers who reach down to the mouthing dog), manage the dog more carefully in those situations until the behavior is more reliable. See teaching incompatible behaviors for the approach to building durable alternatives.