Do productivity apps work? What the research says

The honest answer: sometimes, mostly by accident

A productivity app is software that promises to organize your attention. But do they work? Sometimes, but rarely for the reasons their marketing says. The research on attention suggests that most of what determines your focus has nothing to do with which app you use, and everything to do with how often you are interrupted and how clearly you have decided what to do next.

That is an uncomfortable thing for someone building a productivity app to write. But I would rather start from what the evidence says than from what would be convenient to believe. So here is what researchers have actually measured, where apps genuinely help, and where they quietly make things worse.

What researchers actually measured

The most-cited work here comes from Gloria Mark, a professor at UC Irvine who has spent two decades studying attention in real workplaces. Two of her findings matter most. First, from her interrupted-work studies: after an interruption, people take an average of about 23 minutes to fully return to the original task, not because they are lazy, but because the mind keeps part of its grip on whatever interrupted it. Second, from her 2023 book Attention Span: the average person now spends roughly 47 seconds on any one screen before switching to another.

A related idea, named by researcher Sophie Leroy, is attention residue: when you switch tasks, part of your attention stays stuck on the previous one. Writing a message while half-thinking about a meeting means doing both badly. Leroy's studies showed this residue measurably hurts performance on the task you switched to.

None of this is about apps. It is about interruption and switching. Which is exactly the lens through which productivity tools should be judged, because a tool that reduces switching helps, and a tool that adds switching hurts, regardless of how beautiful its dashboard is.

Why switching is so expensive

Put the two findings together and a workday comes into focus. If screens change every 47 seconds and each genuine interruption costs minutes of recovery, then the bottleneck of modern knowledge work is not effort or intelligence: it is recovery time, paid over and over, all day. Attention residue is the mechanism: the mind does not context-switch cleanly, it smears.

This is also why 'quickly checking' something is never quick. The check takes ten seconds; the return journey takes minutes. A day with four real interruptions does not lose four minutes; it loses the better part of an hour, invisibly.

Where apps genuinely help

The research points to three honest mechanisms. First, offloading: writing something down frees working memory from holding it. A capture habit works in a paper notebook just as well as in an app; the app only wins on retrieval.

Second, commitment: deciding in advance what you will do, and when, is one of the most reliable findings in behavioral science. That is also the core idea behind one intention per day: Peter Gollwitzer's work on implementation intentions showed that a simple 'when X, I will Y' plan dramatically raises follow-through. A tool that forces a real decision once a day is doing something valuable.

Third, scaffolding: a visible timer or a defined work block gives your attention a fence to lean against.

Notice what is missing from that list: streaks, leaderboards, scores, feeds. None of the attention literature suggests that being graded motivates deep work. Those features exist to serve the app's retention metrics, not your focus.

There is one more honest entry on the help side: awareness itself. Simply seeing, once, that your afternoon dissolved into forty screen switches can change behavior more than any feature. Mark's own research found that people who were shown their attention patterns often self-corrected. The value is in the mirror, not the scoreboard: the difference between being shown your behavior once and being ranked on it forever.

Where apps quietly make things worse

The most common failure mode is the tool becoming another interrupter. Every badge, reminder, and 'you haven't logged today' notification is an interruption wearing a helpful costume, and we now know what interruptions cost. The second failure mode is metricization: once your day becomes a chart, it is easy to start performing for the chart instead of doing the work. You optimize the number, and the number stops meaning anything.

There is a subtler cost, too. A heavy tool imports its own maintenance burden: tagging, sorting, reviewing, configuring. The system that was meant to hold your work becomes another workstream. If your productivity system needs its own productivity system, something has gone wrong.

What to look for in a tool

If the research has a practical upshot, it is a short checklist. Choose tools that interrupt you rarely or never. Choose tools that force one clear decision, what matters today, rather than infinite organization. Choose tools that keep your data on your device where you can leave with it. And be suspicious of any tool whose business model depends on you opening it more often than the work requires.

The deepest rule is the simplest: the best tool asks the least of your attention. Attention is the thing being spent; the tool is supposed to be the envelope, not another purchase.

Why I am building one anyway

So why build Focus, knowing all this? Because the bar the research sets is actually clear: decide one thing, protect the time for it, close the day, and never become the interruption. Most tools fail that bar not because it is hard to meet, but because meeting it is bad for their metrics.

Focus is my attempt to build a tool that passes: one intention in the morning, calm blocks through the day, a short reflection at night, and your data staying on your device. Whether it works is for you to judge, but at least the exam it is taking was written by the research, not by an engagement team.