Break Up With Your Phone: Does Catherine Price's Plan Work?

Catherine Price's How to Break Up With Your Phone is the best-selling book in this category and the closest thing it has to a standard programme. A revised edition was published on 4 February 2025, running to 240 pages, with expanded material on algorithms and a new section on children. There is a companion workbook, a free starter kit and a course waitlist on the author's own site.

It is structured as a 30-day plan across four weeks, beginning with a week Price calls Technology Triage: a fact-finding phase before any changes are made.

The publisher describes the plan as evidence-based. That phrase is doing a specific job, and it is worth being precise about what it means, because the answer changes how you should use the book.

No trial has tested this plan

Searches of the literature turn up no randomised or controlled study of Price's 30-day programme. Not a weak one, not a small one. The publisher does not point to one, and the author does not claim one.

"Evidence-based" here means the book is built on research about how phones work on attention and mood. That research is real and much of it is good. What has not happened is the other thing the phrase can imply, which is that this particular sequence of thirty days was given to one group and withheld from another and then measured.

This is not a criticism aimed only at Price. It is true of essentially every popular phone programme, every 30-day challenge and every app's onboarding sequence. Almost nothing in this genre has been tested as a protocol. It is worth saying out loud because readers reasonably assume that a plan marketed as evidence-based has been through something, and the something usually does not exist.

It also does not mean the plan does not work. It means the honest way to assess it is to take it apart and grade the pieces, which is what follows.

The opening week is the control condition

The plan opens with observation: track your use, notice what you reach for and when, take stock before changing anything.

This is intuitively right and it is the part with the weakest support as an intervention. In the best-controlled test of a bundle of phone changes, the control group did exactly this — they monitored their own screen time and nothing else. The intervention group, who made ten small changes, came out lower on problematic use and screen time and better on sleep quality. Self-monitoring was the thing the real intervention had to beat, and it was beaten.

So treat that opening week as diagnosis rather than treatment. It tells you which app to aim at, which is genuinely useful, and tracking a baseline week is worth doing. It is not itself the change. A reader who completes it and stops has, by the evidence, done the placebo arm.

The settings changes are the strongest part

The plan's middle section — notifications, app removal, changing what the phone offers when you pick it up — is where the support is best, and it is the part readers are most likely to skim because it sounds administrative.

The McGill trial above tested precisely this kind of bundle: ten small changes including disabling non-essential notifications and switching the display to greyscale. Effects held for at least six weeks. It is the best effort-to-effect ratio in this entire literature, and it is free.

One correction the book's genre generally needs. The standard instruction is to turn notifications off. A randomised field experiment with 237 people tested three conditions, and the winner was notifications batched into three deliveries a day: more attentive, more productive, better mood, lower stress. The group that received no notifications at all got few of those benefits and reported higher anxiety and more fear of missing out. If you follow one instruction from the whole 30 days, make it the batched version rather than the silent one.

On greyscale, the evidence is real but smaller than usually claimed, and it was one of ten changes in the bundle rather than a tested intervention on its own.

The trial separation is the one with a real study behind it

The plan's best-known element is a weekend away from the phone — a trial separation, followed by reintroduction.

The nearest controlled evidence is a good deal more demanding than a weekend. A randomised trial of 467 adults blocked all mobile internet on participants' own phones for two weeks, leaving calls and texts working. In the group that blocked first, daily screen time fell from 314 minutes to 161, mental health improved by an effect size of 0.56, and objectively measured sustained attention improved by an amount the authors compare to reversing a decade of age-related decline.

Two details from that trial should shape how you read the book's version.

The first is adherence. Of 467 volunteers, 119 met the compliance threshold of keeping the block on for ten of fourteen days — 25.5 per cent, in a sample who had signed up specifically because they wanted to use their phones less. The authors concluded that less restrictive approaches may be beneficial. A 48-hour separation is the less restrictive approach, and there is a reasonable case that its realism is worth more than the larger effect nobody completes.

The second is what happened afterwards. Screen time went back up: 265 minutes a day, against the 314 they started at. The separation does not hold on its own. Which is, to be fair to Price, exactly why her plan is thirty days and ends with a reintroduction phase rather than a weekend and a congratulation. On the broader question, the digital detox evidence is mixed enough that a weekend alone should not be expected to do much.

One thing the book's tradition gets wrong

Somewhere in every programme of this kind is the instruction to keep the phone in another room, usually justified by the idea that its mere presence is consuming your mental bandwidth.

That claim comes from a 2017 study by Ward and colleagues, and it has not held. A 2020 replication using the same tasks found no difference at all between phone-location conditions. The meta-analysis pooling 56 studies and 7,093 participants found a negative effect on working memory, null results across every other cognitive domain, and an overall effect not statistically distinguishable from zero — likely equivalent to zero against conservative bounds.

Keep doing it. Do it because you cannot pick up a phone that is in the kitchen, which is the same access logic behind every other item here. Do not do it because a switched-off phone on your desk is draining your brain, because the pooled evidence does not support that at the strength it is stated.

What the plan gets structurally right

Three things, and they are not small.

It is a bundle. Every trial with a decent result in this field tested a bundle, not a single tip. The McGill study gave people ten changes. The blocking study removed a whole capability. Single-tip advice is what fails, and a 30-day sequence is a bundle by construction.

It is time-boxed. Thirty days is finishable. The one-week version of this — 278 employees asked to cut an hour a day for a single week — produced better mental health, work-life balance and work satisfaction at a two-week follow-up. Short and concrete beats indefinite and vague.

It ends with reintroduction, not abstinence. Given that screen time rebounded to 265 minutes after the two-week block, a plan whose final phase is about what you let back in is addressing the actual failure point rather than pretending it away.

What no plan can do

Everything in this article was measured over two to six weeks. The blocking trial ran four. The friction study, which tracked 1,039 people for an average of 13.4 weeks, is the only one long enough to see what happens next, and what it saw was decay: the proportion of app openings the friction managed to stop declined steadily, and 27 per cent of users took breaks from it.

A 30-day plan ends on day 30. That is the honest limit of the format, and no book resolves it.

Should you read it

Yes, with a clear idea of what you are buying.

You are buying structure and sequencing: a bundle of sensible changes, ordered, time-boxed, with a workbook and a reason to keep going past day four. The genre's evidence says bundles work and single tips do not, so the format is the right one. You are not buying a tested protocol, and the phrase evidence-based should be read as "informed by research" rather than "measured against a control group."

If you want the short version for free: do the audit in an evening rather than seven days, make the settings changes, batch your notifications into three deliveries a day, pick the single app that your own baseline week identifies, and plan what happens in week five, which is the part every study in this article stops before.

For the famous n=1, the New York Times journalist Kevin Roose spent thirty days on a version of this plan with Price as his coach, starting from about six hours a day and more than a hundred pickups, and wrote it up in February 2019. It is a good read and it is one person, unblinded, writing about an intervention he had committed publicly to completing. That is the level of evidence the plan itself currently rests on.