Some trainers differentiate between training the pet dog and training the performance animal-the dog used for search and rescue, police work, hunting and field trials, or as a service dog for the disabled. The view seems to be that the performance dog must be trained with special methods, methods the pet doesn't need because it is only a pet.
This argument is usually put forth by traditional trainers to justify their training methods, especially reliance on correction. That attitude is understandable in someone who has not had the experience of training complex behavior with reinforcers and a marker signal; until you really do it, you can't visualize how it might work.
However, I have seen the same distinction between pets and working dogs drawn by operant trainers. For example, I have heard people say that you can or maybe should skip the clicker and just use food to reinforce some cute behavior with a pet dog; that the clicker should be reserved for the kinds of highly specific responses you would need from a performance dog. I have even heard it said that with a performance dog you should "save" your clicker for particularly difficult situations where the animal has to guess what is wanted (now there's a great way to take the fun out of the clicker).
In practice, however, there's no difference in methodology between clicker training a pet dog and clicker training a performance dog. The clicker works the same way for both. Both learn faster with a clear-cut marker than without. Both can benefit, for example, from the information in a single click. I once heard a police officer describe his dog struggling to locate a thrown object that had come to rest in a spot above the dog's head. When the dog happened to look upward, he marked the behavior, and the dog immediately stood on his hind legs, and was able to see and retrieve the target. In Click for Joy, by Melissa Alexander, a pet owner describes trying for two months to teach her dog to roll over, by saying "Roll over," and luring or helping the dog through the movement. Finally one day she clicked at the moment the dog's weight shifted to the other side; the dog caught on instantly and could roll over on cue from then on.
The real difference between training a pet dog and training a performance dog, in fact, is not the quality of the training, but the quantity. It takes time and effort to teach a dog two hundred distinct behaviors and cues instead of a couple of dozen, and to teach a dog to respond to those cues in any and every circumstance (and that, of course, is true no matter what training approach you use).
Pet owners don't usually want or need to put in that quantity of time. I would not expect my little poodle, Misha, to respond reliably to a sit-stay cue in a burning building. I would not expect Misha to heel reliably through a street riot. I personally know working clicker dogs that do such things; but I sincerely hope that neither Misha nor I will ever be in a burning building or a street riot. So I'm not going to train for it.
If I do feel like training, I teach Misha new tricks. But because Misha is a pet, should I teach him carelessly, or without utilizing the power of the click? Of course not. He is not a second-class canine, deserving only of second-best or inaccurate communication and reinforcement procedures. He needs a smaller repertoire with fewer criteria than a working dog; but he benefits just as much from the experience of being a clickerwise dog. Misha may not get the most training I can do, but he gets the best training I can do. Just as if he were a dolphin.