Argus · Training
Choosing and using rewards: treats, toys, praise, and life rewards
A reward is anything your dog wants enough to work for — and your dog decides what counts, not you. Match the reward's value to how hard the task is: kibble for easy reps at home, real meat or cheese for recall at the park. Keep treats tiny, count them toward daily food, and once a behavior is reliable, move from rewarding every rep to rewarding some reps and using everyday "life rewards" like doors, walks, and play.
At a glance
- Build a treat hierarchy: low, medium, and high value, and use the lowest value that still works.
- Training treats should be pea-sized or smaller; treats should make up no more than about 10% of daily calories.
- Toys, play, praise, and access to things the dog wants are all rewards — use them.
- Fade the food lure early, but don't stop rewarding altogether. Behaviors that are never reinforced fade.
Your dog decides what counts as a reward
A pat on the head might be lovely to one dog and something another dog puts up with. A dry biscuit might be worth a sit in the kitchen but worth nothing when there is a squirrel in view. The test of a reward is simple: does the behavior it follows become more likely? If not, it is not a reward for that dog in that moment.
A quick way to find out: put three different treats on a plate out of reach, let your dog sniff, and see which it goes for first. Repeat a few times. You now know your top-tier reward.
Build a treat hierarchy
| Tier | Examples | Use it for |
|---|---|---|
| Low value | Kibble from daily ration, dry biscuits broken small | Easy, familiar tasks at home |
| Medium value | Commercial soft training treats, dried fish or liver treats | New behaviors, mild distractions |
| High value | Cooked chicken, cheese, hot dog slices, meat-based rolls | Recall, busy outdoor spaces, hard tasks |
| Jackpot | A handful of high-value treats in a row, or a favourite game | Breakthroughs and exceptional recalls |
Use the lowest-value reward that still gets a keen response. If you use chicken for everything, you have nothing left for the moment your dog chooses you over a deer. And if your dog stops taking food at all, that is information too: the environment is too stressful or exciting for learning right now, so move somewhere easier.
Size, calories, and delivery
- Go tiny. Pea-sized for medium and large dogs, smaller for toy breeds. The dog cares about getting a treat much more than about its size, and small treats let you do more reps.
- Count the calories. A common guideline is that treats should be no more than about 10% of daily calories. On heavy training days, take some kibble out of meals and use it in training.
- Soft beats crunchy. Soft treats are eaten in a second; crunchy ones leave crumbs and slow the session down.
- Deliver it where you want the dog to be. Placing the treat at your left leg builds heel position; tossing it away resets the dog for another recall.
Rewards that are not food
Toys and play: For many dogs a quick game of tug or a thrown ball beats any treat, particularly for high-energy behaviors like recall. Keep a special toy that only comes out for training so it stays valuable.
Praise and touch: Useful, but usually weaker than people expect — especially early on. Pair praise with food so it gains value over time, and learn which kinds of touch your dog actually enjoys (many prefer a chest scratch to a head pat).
Life rewards: Anything your dog wants: the door opening, the leash coming off, a sniff of that lamp post, going to greet a friend. Ask for a quick sit or a moment of eye contact, then give the dog what it wanted as the reward. This is how training becomes part of the day instead of a separate activity.
Fading food without losing the behavior
Two different things get "faded", and people often mix them up. The lure (food in your hand used to guide the dog into position) should go early — within a few sessions. The reward (what the dog gets after getting it right) should be reduced gradually and never disappear entirely.
- Stage 1: reward every correct response while the behavior is new.
- Stage 2: once the dog responds fast and reliably in that setting, reward most reps, and give the best rewards to the fastest ones.
- Stage 3: reward unpredictably — sometimes a treat, sometimes a game, sometimes a life reward.
- Always go back to rewarding every rep when you move to a harder place or add a new distraction.
Sources and further reading
- Marker words and clickers — https://argus.dog/library/training/foundations/marker-words-and-clickers
- Training session structure — https://argus.dog/library/training/foundations/training-session-structure
- Come (recall) — https://argus.dog/library/training/core-obedience/recall