Audit Where Your Inbox Time Actually Goes in One Week

The commonly cited figure of roughly 2.6 hours a day spent on email comes from survey data — specifically the US Bureau of Labor Statistics American Time Use Survey and corroborating Microsoft WorkLab research. That number is useful context but almost useless for your situation, because it is an average across millions of different inboxes, roles, and work styles. A recruiter running eight open roles has a very different time profile from a solo operator who sends fifteen replies a day. Repeating the statistic tells you nothing about where your hours go.

An audit does. A one-week self-study with a simple five-column log gives you your own baseline, broken into the four categories that actually matter for deciding what to do about it: reading, deciding, writing, and chasing. The audit is not complicated. It does not require a time-tracking app or a stopwatch. It requires seven days of honest logging and a fifteen-minute review at the end. This guide walks through exactly how to run it.

Before you start: what are the four categories?#

Most people track email time as a lump: 'I spent two hours in my inbox today.' That is not useful because the four activities inside those two hours have completely different causes, costs, and cures. Separating them is the whole point of the audit.

Reading is the time spent scanning, opening, and evaluating messages to decide whether they need you. Most of the volume is here: automated notifications, newsletters, CC replies, FYIs. This category is almost entirely overhead — you produce nothing from it, you just sift. Deciding is the time spent determining what action a message actually requires: a reply, a delegation, a calendar block, a file-and-forget. This is genuine work, but it is often tangled with reading and hard to pull apart.

Writing is composing the replies that do need you. It includes the blank-page friction of starting from nothing, looking up relevant details, and editing before you send. Chasing is the follow-up overhead: resurfacing things you promised to send, nudging people who have not replied, and re-reading old threads to reconstruct their state. Chasing is usually the smallest category in raw minutes and the most expensive when it breaks down.

What does the five-column log look like?#

Keep the log as simple as you will actually maintain for seven days. A notebook, a sticky note, a plain spreadsheet — the format does not matter. Each row is one email session, and it has five fields.

A session is any continuous block of email work. If you sit down at 9 a.m. and write two replies before switching to a meeting, that is one session. If you check email quickly at 2 p.m. because a notification distracted you, that is a separate session — and note what triggered it, because notifications as a trigger tell you something different than a deliberate batch-check.

Assign one category per session even if you did more than one thing. Use the activity that took the majority of the time in that block. Mixed sessions are fine — just pick the dominant one. Precision here is less important than consistency across seven days.

Running the audit: seven days, step by step#

1. 1

   Choose a representative week

   Pick a week that is ordinary for your role — not a launch week, not a holiday, not a week you are travelling. The audit is only useful if the week reflects your typical workload. If you have two distinct inbox modes (for example, client-facing days and internal-only days), note which kind each day is, so you can read the breakdown in context.

2. 2

   Open the log before you open your inbox each morning

   Place the log where you will see it. A tab pinned next to your email, a paper notebook on your desk, a sticky note on your monitor. The mechanical problem with this audit is not remembering to log at the start of a session — it is remembering to log the end. Build the habit of closing the row before you switch tasks.

3. 3

   Log every session, including the two-minute checks

   Notification checks, reflex glances, quick scans between meetings — log all of them. These short, fragmented sessions are often where a surprising portion of total time hides. A 90-second check that happens twelve times a day is eighteen minutes, and it shows up as twelve rows in the Reading column. That pattern is something you cannot see without the log.

4. 4

   Do not change your behaviour during the audit week

   The point is to capture what actually happens, not what you wish happened. If you normally check email on your phone during lunch, log it. If you reply to an important message at 10 p.m., log it. Auditing a sanitised version of your inbox week gives you a baseline you cannot use.

5. 5

   On Friday afternoon, total each column

   Add up the minutes for each category across all seven days. You want four numbers: total reading minutes, total deciding minutes, total writing minutes, total chasing minutes. Then count the number of sessions in each category. Two numbers per category — total time and session count — tell you more than one.

Platform differences: where to find supporting data#

The log is the primary record, but every major email platform gives you supplementary data you can use to sanity-check your numbers or fill gaps. None of these are a substitute for the log — they measure activity, not time spent on that activity — but they help.

How to read the result without deceiving yourself#

The audit produces four numbers. Here is what each pattern tends to mean and where people misread it.

A large reading total — more than half of your email time spent in the reading category — almost always means triage is broken. You are processing a large volume of messages that are mostly noise, and most of them are pulling you in via notification rather than deliberate batch-check. The session-count column will confirm this: many short Reading sessions is a notification problem; a few long ones is a batching problem.

A large deciding total often means your inbox is full of ambiguous messages that are not clearly actionable. That is sometimes a sender problem (people emailing you unclear requests), sometimes a system problem (nothing has been delegated or filtered), and sometimes a signal that you are doing work in the inbox that belongs somewhere else — a project tracker, a shared doc, a quick call.

A large writing total is straightforward for some roles and surprising for others. Founders and salespeople expect it; managers sometimes do not. If writing is your biggest category and most of what you write is routine answers to predictable questions, you have identified a high-leverage target: the replies that are routine but not quite templatable are exactly where drafting assistance cuts the most time.

A large chasing total in minutes is a follow-up discipline problem. But also look at the note column: if chasing rows are concentrated in a few threads or a few senders, the problem may be narrower than it looks. And remember that this category is small in time but large in consequence — one missed chase can cost more than a full week of chasing minutes.

Here is a concrete example of how the numbers pull apart. Suppose a week of logging ends with 4 hours 15 minutes reading across 47 sessions, 1 hour 40 minutes deciding across 22 sessions, 2 hours 5 minutes writing across 14 sessions, and 55 minutes chasing across 9 sessions — 8 hours 55 minutes across five working days. The reading total is not the real surprise; the session count is. Forty-seven reading sessions in a week averages nine pulls per day, most under three minutes and most triggered by notification banners for messages that did not need to be read in real time.

Now look at writing. Two hours five minutes across fourteen sessions sounds modest, but if eight of those fourteen sessions were replies to variations of the same three stakeholder questions — clarifications about deadlines, status, and scope, rewritten every time — the audit has surfaced a specific target: snippets for the three recurring answers would cut the writing total by roughly a third. Two clear actions come out of the numbers by Friday afternoon: silence the notification sources dragging reading into 47 fragmented sessions, and template the top three replies. Neither would have surfaced from a lump total, and neither is a general productivity tip — both are specific to what this log actually showed.

What to do when the log breaks down#

Most people run into at least one of three problems during the audit week. Here is how to handle each without abandoning the method.

• You forgot to log several sessions. Estimate them from memory at the end of the day, mark them as estimates, and include them anyway. A slightly imprecise log is more useful than a log with gaps you left empty to stay 'honest.' Your memory of the category and rough duration is accurate enough for this exercise.
• Your sessions blurred together — you went from reading to replying to chasing without a clear break. Split the session at the point where the dominant activity changed and log two rows with estimated timestamps. This is normal for the first two days; the act of logging tends to make session boundaries clearer as the week goes on.
• You were unusually busy or unusually quiet. Note it. If the week was a genuine outlier, run the audit a second week. But do not let the fear of an unrepresentative week stop you from completing the first one — an outlier with good notes is more useful than no data at all, and the category breakdown is usually more stable across unusual weeks than the total time is.
• You started editing your inbox during the audit — unsubscribing, filtering, muting threads — because the log made the noise so obvious you could not sit with it. This is common by day three and it corrupts the baseline you were trying to capture. Finish the week logging without changes, then run every clean-up idea you had during week two and log that week too. The comparison between the two weeks is the highest-value artefact the whole exercise produces, and you only get it if you resist the urge to fix things mid-audit.

A faster way to track this continuously#

A one-week manual audit gives you a baseline. What it cannot do is update that baseline as your inbox changes — when you take on a new role, change how you handle follow-ups, or delegate a category you used to own. Running the full audit again from scratch every quarter is realistic for some people and not for others.

We build AI Emaily, an AI-native email client that tracks the structure of your inbox activity as a byproduct of how it works. Because it reads, categorises, and handles incoming mail on your behalf, it has a continuous view of your volume by category — what is arriving, what is routine, what requires a reply, what is waiting on someone else — that surfaces in the living brief rather than needing a manual log. If you want to see where your inbox time goes without logging sessions by hand, that is the practical alternative. Autopilot and Copilot modes mean you can start with full human approval and graduate to autonomous handling only for the categories you choose, so the audit insight translates directly into targeted automation rather than a general guess. Start with a 7-day free trial at app.aiemaily.com/signup, or see the pricing at /pricing.