Okay, But Do These Buttons Actually Work?
Before I told you to spend money on any of this, I wanted real answers — not just cute videos. So I went and read the actual research. Here's what's real, and what's still an open question.
I'll be honest: when talking buttons first crossed my feed, my gut reaction was "that's adorable, and also probably wishful thinking." Dogs are experts at reading us — not exactly experts at grammar. So before recommending you buy anything, I dug through the peer-reviewed research instead of just the viral clips. Turns out there's more real science here than I expected, and it comes with some honest limits too.
Dogs Understand What the Words Mean
Researchers at UC San Diego ran two experiments — one in person with 30 dogs, one as a much larger at-home citizen-science study using dogs their own owners had already trained. In the in-person version, someone pressed a button (or just said the word out loud) and then watched what the dog actually did next. Press Play, and dogs showed roughly seven times more play-related behavior than usual. Same story with Outside.
Here's the part that got me: dogs responded the same way whether a total stranger pressed the button or their own owner did, and the same whether the word was spoken out loud or played from the button itself. That rules out the easiest explanation — that dogs are just reading their owner's face or body language, not the actual word. (Food didn't show a clear effect, but the researchers think that's just because the dogs weren't hungry during testing, not because the idea failed.)
Read the full study on PLOS ONE →They Might Even Be Combining Words
A follow-up study looked at what happens when dogs press two buttons back-to-back — something like Outside then Want. Those two-button combos weren't random. Dogs produced certain pairings far more often than chance would predict, and more often than you'd expect if they were just copying whatever combo their owner happens to press.
It's a long way from "dogs have grammar." But it's a real, measurable pattern — not just adorable noise.
Read the full study in Scientific Reports →What This Doesn't Prove (Yet)
Here's where I want to be straight with you instead of overselling it: understanding what a word predicts is not the same thing as a dog deliberately "talking" to you the way that word implies. Even the researchers running these studies are careful about that distinction — whether a dog is intentionally picking the right button to ask for something, versus just having learned "this button happening means this thing happening," is still an open question they're actively studying.
There's also a classic trap in this kind of research called the Clever Hans effect, named after a horse that seemed to solve math problems on cue but was actually just reading tiny, unconscious signals from his handler. The newer studies specifically design around this risk — strangers pressing the buttons, blind video review — which is reassuring. But it's exactly why you'll see actual researchers, and outlets covering this research, describe it as "real comprehension, still-open question about intent," rather than "dogs can talk, case closed."
Read the AKC's take on the science →So... Is It Worth Trying?
Personally? I landed on yes. Even with the open questions, there's real, peer-reviewed evidence that dogs — and probably cats too, though most research so far has focused on dogs — understand these words and respond to them consistently. That's more than enough for me to think it's worth the pretty low cost of trying: a button, a little patience, and about two weeks of feeling slightly ridiculous narrating your life out loud.
Worst case, you've taught your dog three new words. Best case, you've got a genuinely clearer window into what they're thinking. Either way, nobody's worse off.
Ready to see for yourself?
The training method that actually worked for us is laid out step by step — no science background required, just three buttons and some patience.