Do Email Folders Actually Work? What 345 Inboxes Revealed

The most cited research on email folders concluded the opposite of what most people assume. In a 345-person field study at IBM Research, Steve Whittaker and his team instrumented a real email client and watched users perform more than 85,000 retrievals over several months. Searching for an email took about 17 seconds on average. Finding the same email by clicking through folders took 58 seconds. The people with the most elaborate folder hierarchies were not better at finding what they needed; they were just spending more time on the upkeep.

So no, email folders mostly do not work. They feel like organization but mostly produce the illusion of organization, and the illusion costs you time on both ends: filing each message, then later digging through the structure you built.

This is a contrarian post, but the data is unusually clear. Email is one of the rare domains where the research finding directly contradicts the common-sense intuition. The intuition says "more structure = easier retrieval." The data says structure has a cost, and search ate the benefit roughly fifteen years ago.

The deeper problem is that filing email is not just slow; it's a categorization task disguised as a productivity ritual. Every incoming message becomes a tiny decision: Project A or Project B? Receipts or Finance? "Important" or "Reference"? Multiply that by 121 emails a day per Radicati Group's 2023-2027 Email Statistics Report and you've built yourself a second job sorting messages you may never reopen.

Our stance, for context: the problem is not that you're filing email wrong. The problem is that an inbox without a finish line will never feel finished, no matter how clever your folder tree. More on that later.

Contents#

• What the IBM Research Study Actually Found
• Why Folders Feel Productive (Even When They Aren't)
• Are Folders Ever Worth Building?
• What to Do Instead: A Minimal System
• The Folder Trap and the Finish Line
• Frequently Asked Questions
• Sources

What the IBM Research Study Actually Found#

The canonical study here is Am I Wasting My Time Organizing Email? A Study of Email Refinding, published at CHI 2011 by Steve Whittaker, Tara Matthews, Julian Cerruti, Hernan Badenes, and John Tang of IBM Research–Almaden. The team logged real email use across 345 long-term users of an enterprise client that supported folders, tags, threading, and search side by side.

Three findings did most of the damage to the pro-folder argument.

Search was 3.4x faster than navigation. Refinding via search took a median of 17 seconds. Refinding by clicking through folders took 58 seconds. That gap held even when participants knew where they had filed something.

Filers weren't more successful; just slower. Heavy filers (people who created many folders and used them) had no measurable advantage in finding success rates compared to people who barely organized at all. They simply spent more time preparing.

Most retrieval wasn't via folders anyway. Opportunistic behaviors; scrolling the inbox, threading, and search; accounted for roughly 87% of refinding actions, even among committed filers. People built folder structures and then mostly bypassed them.

The original Whittaker and Sidner CHI 1996 paper that coined "email overload" had already hinted at this. A 2014 follow-up study by Fisher, Brush, and Whittaker revisited the findings a decade after powerful search arrived in mainstream clients and confirmed that the case for elaborate filing had collapsed once Gmail-style search became universally good.

One caveat worth naming. The IBM study used an enterprise email client circa 2011. Search has only gotten better since; Gmail, Outlook, and modern iOS clients all index attachments and inline images now, which the 2011 systems mostly didn't. The gap has almost certainly widened.

Why Do Email Folders Feel Productive When They Don't Work?#

Filing feels good because it is a clean, visible action with an immediate payoff: a message disappears from view. The dopamine of "I dealt with that" arrives at the moment of filing, not at the moment of later retrieval.

The retrieval cost, by contrast, is invisible at the time of filing. You have no idea, when you're dragging an email to the Q3 Vendor Contracts folder, that future-you is going to spend two minutes hunting for it across Vendors, Contracts, Q3, and Procurement because past-you was inconsistent.

Cognitive scientist Ofer Bergman, who has studied personal information management for two decades, has documented this pattern across email and file systems. In The Effect of Folder Structure on Personal File Navigation (Bergman et al., 2010), he and his collaborators found that even when search engines were available, users navigated to files via folders 56–68% of the time and only searched as a last resort; but their navigation success was tightly bound to keeping folders shallow (fewer than 3 levels deep) and small (around 12 items). Once folders grew past those thresholds, navigation collapsed and people fell back to search anyway.

"Filers expend significant effort organizing their email in advance, but this preparation does not improve retrieval success."; Whittaker, Matthews, Cerruti, Badenes, and Tang, IBM Research, CHI 2011

There's also a mood effect that makes the problem self-reinforcing. Research on personal information categorization, summarized in Oh's 2012 study at the American Society for Information Science, found that people in negative or stressed moods create more folders containing fewer items at deeper nesting; the exact structure that the Whittaker data shows hurts retrieval. The folder system you built during a stressful Tuesday may be the one actively slowing you down on a calm Thursday.

Are Folders Ever Worth Building?#

Yes, but the cases are narrower than most people think. Folders earn their keep when the structure maps to a real-world legal, audit, or compliance boundary that future-you (or someone else) will need to defend.

Specifically:

• Tax and financial records. A Tax 2026 folder is genuinely useful because the audit boundary is the calendar year, the retention rule is statutory, and search alone won't reliably surface a receipt you didn't know to look for.
• Active project archives that span multiple collaborators. If three people will need to refind the same thread for the same project, a shared label cuts the search cost across all of them.
• Legal hold and discovery. This is non-negotiable in regulated industries; but it's a compliance requirement, not a productivity feature.
• Travel and shipping confirmations. A Travel label that auto-applies via filter (so you do zero filing work) lets you scope a search to "the United Airlines email I got in March" without building a parallel structure.

Notice the pattern: the cases that justify folders are ones where (a) the structure is enforced by an external rule, not your judgment, and (b) the filing happens automatically via filters, not manually. The moment you're hand-sorting Important vs Maybe Important vs Probably Junk But I'm Not Sure, you have rebuilt the inbox inside a folder and gained nothing.

The 6-folder system that productivity blogs love (Action / Waiting / Reference / Archive / Read Later / Done) sounds reasonable on paper. In practice, people abandon it within weeks because the categories overlap and the filing decision is harder than the response itself. Is this Action or Waiting?; and now you have a second decision queue.

What to Do Instead: A Minimal System#

The research-backed answer is anti-climactic: one archive folder, aggressive search, and filters for things a machine can categorize without your help.

A workable setup looks like this:

Source: synthesized from Whittaker et al. (2011) and Bergman et al. (2010) recommendations.

That's the whole system. The reason it works is that it offloads categorization to the search algorithm, which is genuinely better at it than you are, and it removes the daily decision tax of "where does this go?"

If you currently have 40 folders, the migration is easier than you'd guess. Select all in each folder, move to Archive, delete the folder. You will lose nothing; search indexes the entire archive; and you will reclaim the cognitive overhead of maintaining the tree.

Honestly, the part I find hardest is psychological. A folder tree feels like control. Trusting search feels like letting go. The first week without folders is uncomfortable in a way the data doesn't capture.

The Folder Trap and the Finish Line#

Here is the deeper issue, and it's the one folder evangelists rarely address. Folders don't reduce the size of your inbox. They redistribute it.

You can have an immaculate 12-folder system and still be drowning. Filing isn't finishing. A message moved from Inbox to Reference is still mentally open; you've just hidden it from view, which is precisely the move that the Zeigarnik effect research shows keeps unfinished tasks burning attention in the background.

This is the same problem that makes inbox zero unstable as a goal. You can hit zero, file the last message, and the inbox refills five minutes later. The folder structure doesn't slow that down. It just gives you something to do while it happens.

Cal Newport, Georgetown computer science professor and author of A World Without Email (Portfolio, 2021), makes a related argument: the productivity ritual around email; folders, labels, color-coded flags; is a coping mechanism for an unbounded medium. Newport's frame is that email lacks "structured workflows," but the folder critique is downstream of his point. If the medium has no finish line, no amount of internal organization fixes it.

Swizero's design is the opposite of a folder tree. There is no folder tree. The inbox is a fixed card limit; a handful of cards, processed with one-handed swipes, and when they're done, you are done. The "where does this go?" question is replaced with "what does this need?"; reply, archive, delete. The categorization decision evaporates because there's nowhere to file to.

That's not a folder system. It's the absence of one, by design.

Frequently Asked Questions#

Is filing email actually a waste of time?#

For most personal inboxes, yes. The 2011 IBM Research field study of 345 users found that people who filed extensively were no better at refinding emails than people who didn't file at all; and they spent significantly more time on inbox upkeep. The exceptions are tax/financial records, legal compliance, and shared project labels with multiple collaborators.

Why is search faster than folders?#

Search retrieves directly from a query in your head ("flight confirmation United March") to the message. Folder navigation requires you to first remember the structure you built, then make several click decisions, then visually scan the folder contents. Whittaker's IBM study clocked search at about 17 seconds versus 58 seconds for folder navigation, and the gap has only grown as Gmail- and Outlook-style indexing has improved.

How many folders should I have?#

If you're going to use folders at all, the research suggests fewer than 5–7, kept shallow (no nesting beyond two levels), with as much filing as possible automated via filters. Bergman et al. (2010) found navigation success collapses once folders grow deeper than 3 levels or contain more than ~12 items each.

Should I delete all my email folders?#

Probably most of them. A safe approach: keep one Archive folder, keep filters that auto-sort newsletters and receipts, and delete folder hierarchies you built manually. You won't lose any messages; they'll all stay searchable. Try it for two weeks before committing; the discomfort is mostly psychological.

Are Gmail labels different from folders?#

Functionally similar, but labels can stack (one email gets multiple labels), so they're slightly more flexible than traditional folders. The research doesn't show that this flexibility translates into better retrieval; it just adds more decisions per message. The same minimal-system advice applies.

Sources#

1. Am I Wasting My Time Organizing Email? A Study of Email Refinding. Whittaker, Matthews, Cerruti, Badenes, and Tang, IBM Research–Almaden, CHI 2011. The 345-user field study comparing search and folder navigation. Search: 17s. Navigation: 58s.
2. The Effect of Folder Structure on Personal File Navigation. Bergman, Whittaker, Sanderson, Mizrachi, Beyth-Marom, and Henderson, Journal of the American Society for Information Science and Technology, 2010. Navigation success depends on shallow, small folders.
3. Revisiting Whittaker & Sidner's "Email Overload" Ten Years Later. Fisher, Brush, Hogan, Smith, and Jacobs, CHI 2006. Confirms the search-over-filing finding holds in modern clients.
4. What Happens Once You Categorize Files Into Folders?. Oh, Proceedings of the American Society for Information Science and Technology, 2012. Mood-dependent folder creation patterns.
5. Email Statistics Report 2023-2027. Radicati Group, 2023. 121 business emails sent and received per user per day.
6. A World Without Email. Cal Newport, Portfolio, 2021. The structured-workflows argument for why email's unboundedness is the root problem.