Why Phone Notifications Feel Like Dog Treats (And How to Stop Rewarding Bad Behavior)

If you’ve ever trained a dog, you know how powerful intermittent reinforcement is. You give the treat sometimes — not always — and the dog checks in with you constantly because it learned that checking is worth it. Sometimes there’s a treat. Sometimes there isn’t. The not-knowing is what makes it compelling.

Your notification feed works the same way.

The variable reward loop

Behavioural psychologist B.F. Skinner demonstrated in the 1950s that variable ratio reinforcement — reward delivered after an unpredictable number of responses — produces the most persistent behavior. Slot machines, email, and social media feeds were all (intentionally or not) designed around this principle.

When you pick up your phone, sometimes there’s something good. A message from a friend, news about something you care about, a reply you’ve been waiting for. More often there isn’t. But the not-knowing is exactly what makes putting the phone down feel difficult — you’re mid-slot-pull, and the reel hasn’t stopped yet.

This isn’t a character flaw. It’s how the reward circuit in the brain responds to unpredictable outcomes. The same mechanism that kept your ancestors checking bushes for berries is being hijacked by a push notification from a food delivery app.

What apps are actually training you to do

When an app sends a notification, it’s reinforcing a behavior it wants: opening the app. Each ping is a cue that triggers checking. Over weeks and months, the habit becomes strong enough that many people check their phones in the absence of any notification at all — the behavior has become somewhat automatic.

The subtlety is that you’re being trained by systems that are explicitly optimized for engagement. The timing of notifications, the wording, the red badge count — these are not accidents. They’re A/B-tested stimuli in a behavioral loop that benefits the app’s metrics, not your attention.

What dogs can teach us about setting limits

Here’s the useful part of the pet analogy: we actually know how to manage this.

Don’t reinforce checking. If you check your phone every time you feel the urge, you’re reinforcing the urge. The way to weaken a conditioned behavior is to not respond to the trigger. Batch-processing notifications (checking at set times rather than every ping) breaks the loop gradually.

You control what triggers the check. A well-trained dog doesn’t beg at the table because begging was never rewarded there. Apps don’t get to demand your attention on surfaces or at times where you’ve established that checking doesn’t happen. Do-not-disturb schedules, app timers, and locked focus modes are ways to enforce this spatially and temporally.

Eliminate the unpredictability. The variable reward is what makes the loop compelling. If you consistently find that checking is unrewarding, the urge weakens. Unsubscribing from low-value notifications removes the “treat” from most of the pings, which eventually changes what checking feels like.

Separate high-signal from low-signal alerts. Not all notifications are the same. A text from a family member is worth interrupting yourself for. A badge count on a shopping app is not. The goal isn’t zero notifications — it’s knowing which alerts deserve to interrupt you, and setting up your phone so that the ones that don’t, can’t.

The practical starting point

You don’t need to do all of this at once. A single change tends to be more effective than an overhaul anyway, because you can actually observe what it does.

Pick the notification type that interrupts you most often without adding value — usually social media badges, promotional emails, or news apps — and turn it off for two weeks. Not silent. Off. The absence of the cue is what allows the checking reflex to weaken.

Most people who do this find that after two weeks, the habit around those apps shifts noticeably. The automatic checking reduces, the urge to open the app becomes more deliberate, and the sense of missing out fades because you learn quickly that you weren’t missing much.

Training goes both directions. You can train your apps to reach you only when they’ve earned it.