Data Driven Dogs
Behavior Modification
Jumping on People
You know the routine: every time someone walks through the door, your dog launches up, and you’ve said "down" or "off" more times than you can count. This isn’t rudeness. It’s a behavior that’s worked—reliably—for attention from humans, every time. Jumping works. That’s the problem, and it’s also the solution.
Define What Your Dog Will Do
Not "stop jumping." That’s just an absence, and you can’t reinforce an absence. The target behavior is four paws on the floor during greetings. Or better: sit when a person approaches. That’s real, observable, and reinforceable. Once you know what you’re building, you stop reacting to what you don’t want and start teaching what you do.
Why This Behavior Persists
Jumping sticks because it’s powered by attention. Any attention. Pushing your dog off is attention. Saying "no" is attention. Even eye contact while backing up is attention. Negative reactions still give your dog the social contact they’re after. The reinforcement schedule is rich, immediate, and delivered by every human the dog meets.
The matching law explains why your dog "won’t listen": the competing reinforcement for jumping (instant attention from excited humans) is richer than what’s been offered for sitting (occasional treat, sometimes). You don’t need to punish the jump. You need to tip the ratio.
The Training
1
Become boring when the dog jumps
When your dog jumps, turn your back completely. Cross your arms. Look at the ceiling. Become the least interesting object in the room. No eye contact, no words, no pushing. The instant all four paws touch the floor, turn back and deliver calm attention. Your timing matters: paws hit the floor, you turn around within 1 second. Wait 3 seconds and you’ve missed it.
2
Ask for the sit before the jump happens
Once your dog is quicker to return to the floor, start cueing the sit before contact. You see the approach, momentum building—say "sit" before the front feet leave the ground. The sit earns a treat plus the greeting. If the dog approaches without sitting, you turn away. You’re building a new habit loop: person approaches → sit → good things happen. This is classic Differential Reinforcement of Alternative Behavior (DRA).
3
Practice at the front door with a helper
Doors are where jumping peaks because excitement peaks. Have someone ring the doorbell or knock. Cue your dog to sit before you open the door. If the sit breaks, the door closes. Same logic as door charging: the thing the dog wants (the visitor) is gated behind the behavior you’re building. When the sit holds, the door opens and the visitor greets your dog calmly. This is an example of using natural reinforcers to strengthen the new behavior.
4
Recruit your visitors
The hardest part isn’t your dog. It’s other people. Visitors who say "oh, I don’t mind!" are accidentally maintaining the behavior. A direct request works: "Can you help me train my dog by only greeting her when she’s sitting?" People respond well when they feel like they’re helping, not being corrected.
Praise the 99.9%
Dunbar makes this point sharply: most owners only give feedback when something goes wrong. The dog lies quietly for three hours and gets nothing. The dog jumps once and gets a reaction. That pattern teaches the dog that misbehavior is the only way to get your attention.
Flip it. When your dog sits near a guest without jumping, say "good dog." When your dog approaches and keeps four on the floor, acknowledge it. You’re building a reinforcement history for calm behavior that competes with the history for jumping. Over time, calm becomes the default because calm is what pays.
The Long Game
This training takes consistency from every person the dog encounters, and that’s the hardest part. But the natural reinforcer is built in: polite dogs get more of what they want. More greetings, more off-leash freedom, more social access. The sit isn’t a trick. It’s a passport.
Grounded in Susan Friedman's functional behavior framework, differential reinforcement principles (DRI), Ian Dunbar's representative feedback methodology, and Jean Donaldson's greeting protocols. Data Driven Dogs, Mercer Island WA.