Amazon may stumble from time to time and also pull down bad publicity, but its instincts and execution are often stunningly good.

We're seeing the latest example in this week's announcement of a new service called Audible for Dogs.

No, you're not reading The Onion. Yes, they're serious. And the idea is brilliant.

On the surface, I admit it will sound nutty to many. Here's Amazon's description: today announced the release of Audible for Dogs, an audio content destination designed to foster calm, relaxed behavior in dogs. Developed in collaboration with Cesar Millan, Audible for Dogs provides dog owners and caregivers peace of mind when leaving their dogs at home alone, knowing the dog will have the comfort of a human voice. Featuring Cesar Millan's Guide to Audiobooks for Dogs, an original piece of content written and narrated by Millan--available free at launch--in addition to a curated rotating selection of audiobook titles with personalized introductions from Millan about their canine appeal, Audible for Dogs is available at

To many people, this is going to seem like an utter scam and waste of time. But one person's ridiculous is another's genius idea. For dog lovers, keeping the pet happy while they're off at work or wherever is important.

Not only do they want the dog to feel happy, the owners also want their pets to remain calm. If you've ever seen a dog without enough exercise or social contact after spending the day in a house or an apartment, then you know what it's like. You can face hyper on a level that a kid riding a sugar high might never hit.

Because they have to make a living, these people are primed to feel guilty, so some emotionally resonant marketing can have a big impact. Guilt is a classic driver of behavior.

But pet owners often leave music or the television on while out so there's something to keep the dog some kind of company. That means no extra money being spent by anyone -- and no added profit for Amazon.

The company's argument is that not all sound is the same. It points to research from Hartpury College in the U.K.:

The study investigated the effects of audiobooks on 31 dogs housed in a rescue centre. They were exposed to five different auditory conditions - audiobooks, classical music, pop music, music 'designed' specifically for dogs and no additional sound for the control group. The dogs' behaviours were recorded every five minutes during a two-hour period in each 'condition'. The duo found that listening to actor Michael York's reading of "The Lion, the Witch and the Wardrobe" relaxed dogs more than pop and even classical music from "The Best of Beethoven".

One study of 31 dogs in a single facility that is nothing like being in a home is far from statistically satisfying. Although the results were potentially interesting, you can't project the results beyond what was observed.

Audible worked with former television personality Milan, who can be controversial in his own right, to undertake its own study with 100 volunteer participants from April 2017 to May 2015. Each participant had to leave the house for at least an hour during a type of audio program and received an Amazon Echo smart speaker to play the material. The people filled out surveys on their dogs each day.

Now, this to me doesn't seem like much of a study either. One hour minimum? Were people out longer, like 8 hours at a time? Did they have to be around much of the day to record canine reactions to different programming? That would make the dynamics different from someone being away from home all day.

How was programming swapped around? According to the write-up, it doesn't even sound as though they played anything but audiobook narration. So even if three out of four people thought the dogs were happier, might something else, like calm music or leaving the television on, have done the same?

Frankly, though, that's unlikely to matter. Few consumers actually dip into the details of research used by companies. What they will hear is that their dogs will be happier and the cost of an Audible account is chicken feed, pardon the cross-species reference, compared to what they likely spend on food, toys, and vet bills. Audible will snag new subscriptions and Amazon will gain that much more data about users, which by itself is worth quite a bit to the company. Plus, an Echo runs about $179, which isn't chump change but still less than a short stay in a kennel.

However, the owners better hope that the Alexa service doesn't start to understand canine requests. Or they could find themselves purchasing more chew toys than anyone thought possible.