Most Slack messages and emails are noise, not signal — and treating every notification as a to-do item is what makes messaging feel like a second job. The fix is a triage funnel: sort inputs into noise (ignore), signal (batch and read), and commitment (extract immediately) — then set your channels, notifications, and cadence to enforce that sort automatically.
Quick answer: Sort every message into noise, signal, or commitment. Ignore noise, batch signal into two or three windows a day, and the moment you spot a commitment, move it out of the chat app entirely — no exceptions.
Treat Messaging As a Triage Funnel, Not a To-Do List
A triage funnel works because most messages don't deserve equal attention. Roughly 80% are noise — FYIs, reaction threads, channels you're CC'd on "just in case." A smaller slice is real signal worth reading, and a tiny fraction are commitments: promises that need to live somewhere permanent, not in a scrollback buffer.
The mistake most PMs make is inverting this. An inbox or #general channel with an unread badge feels like a queue of obligations, so it gets treated like one — read top to bottom, in arrival order, as if the person who happened to type fastest earned your afternoon. That's not prioritization; it's surrender to whoever pinged last.
| Category | What it looks like | What to do with it |
|---|---|---|
| Noise | Broadcast FYIs, emoji-reaction threads, channels you're on "just in case" | Skim or skip — it should never generate a task |
| Signal | Relevant decisions, questions inside your domain, real customer feedback | Read in a scheduled batch; note if it needs a follow-up |
| Commitment | A deadline, a "can you...", a promise you or a teammate made | Extract immediately into a task, calendar block, or reminder |
This isn't a new idea — it's the namesake of Basecamp co-founder Jason Fried's long-running Signal v. Noise blog, which has argued for two decades that most workplace communication tools optimize for volume, not value. Cal Newport makes a related distinction in Deep Work between shallow work (reactive, low-value, easily interrupted) and deep work that actually moves a roadmap forward. Every unfiltered notification is a vote for shallow work, cast by someone else, on your calendar.
The research backs the discomfort you already feel. A 2016 Harvard Business Review study by Rob Cross, Reb Rebele, and Adam Grant found that employees spend upwards of 80% of their workday in collaborative activities — meetings, email, chat — leaving a shrinking sliver for the individual thinking a PM role actually requires. A triage funnel is how you claw part of that sliver back without going dark on your team.
Building this funnel is really one piece of a larger personal PM operating system — the set of defaults that decides what gets your attention before you have to decide in the moment. Slack and email are just the two inputs with the loudest alarms.
Tier Your Slack Channels So Not Every Ping Earns Your Attention
A three-tier channel scheme — must-see, scan-daily, and archive-only — turns an undifferentiated wall of channels into a ranked list your notification settings can actually enforce. Without tiers, Slack treats a channel about the office coffee machine the same as one where your VP asks a direct question, because Slack has no idea which one matters more. You do.
Define the three tiers once, then apply them ruthlessly
- Tier 1 — must-see: Your direct reports, your core project channel, and any thread where you're the named decision-maker. Full notifications, badge and banner both.
- Tier 2 — scan daily: Cross-functional channels you contribute to but don't own. Badge only, no banner — you'll see the count, not get interrupted by it.
- Tier 3 — archive-only: Company-wide announcements, social channels, anything you joined once and never needed again. Muted, checked (if ever) on your own schedule.
| Tier | Example channels | Notification setting | Check cadence |
|---|---|---|---|
| Tier 1 | Your squad channel, direct manager DMs, escalation channel | All new messages, sound on | Real-time to hourly |
| Tier 2 | Cross-functional project channels, design/eng syncs you're looped into | Mentions only | 2-3x per day |
| Tier 3 | Company-wide, social, alumni of old projects | Muted entirely | Weekly skim, or never |
Most PMs never do this sorting because it takes twenty honest minutes, and it's tempting to leave every channel on "everything" so nothing slips through. But that instinct is exactly backward — a tool audit applies to notification settings just as much as it applies to your software stack. If you've never audited which of your fifteen tools you actually use, your Slack sidebar is very likely carrying the same bloat.
Re-tier quarterly. A channel that mattered during a launch can safely drop to Tier 3 once the launch ships; a channel you muted eight months ago might now house the one conversation you actually need. Treat the tier list as a living document, not a one-time setup task.
Prune notifications at the app level, not just the channel level
Channel tiers only work if the notification settings underneath them are equally deliberate. Turn off badge counts on your phone's home screen for Slack and email entirely — a red circle with a three-digit number is a physiological trigger, not useful information. Desktop banners should exist only for Tier 1.
- Disable
@channeland@herenotifications from any group you don't lead — they're a broadcast tool being misused as an urgency tool. - Turn off notifications for reactions (emoji) and thread replies you're not actively part of.
- Set explicit quiet hours in both Slack and your email client, not just "do not disturb" you remember to toggle manually.
- Unsubscribe from digest emails and CC threads where you're not the primary audience — being looped in isn't the same as being responsible.
Build a PM Email System That Doesn't Run Your Calendar
An inbox with a high unread count is not a stack of obligations — it's an unsorted mailroom. A durable pm email system stops treating "0 unread" as the goal, and starts treating "nothing important sat unactioned for 48 hours" as the real one.
Unread counts measure arrival, not importance. A newsletter you'll never read and a stakeholder's budget approval both count as "1 unread" — the number tells you nothing about which one matters. Chasing the number to zero optimizes for the wrong variable entirely.
Unread badges are a queue, not an obligation. You are allowed to leave low-value email unread indefinitely — that's what the badge is for.
A folder structure that mirrors the triage funnel
Set up three working folders (or labels, depending on your client) that map directly onto noise, signal, and commitment:
- Action — anything requiring a reply, a decision, or a document from you this week. Checked first, every batch.
- Reference — decisions, specs, and threads you may need to search later but don't need to act on. Filed, not read top-to-bottom.
- Watching — threads where you're CC'd for visibility only. Skimmed last, if at all.
Rules-based filters can do most of the sorting automatically: automated reports and calendar notifications route straight to Reference, and any email where you're only in the CC line (not the To line) can skip your primary inbox entirely. Most email clients — Gmail, Outlook, Superhuman — support this natively; the work is deciding the rules once, not maintaining them daily.
The Rule That Actually Saves You: Commitments Leave the Chat App Immediately
The single highest-leverage habit in this whole system is a hard rule: the moment a message contains a commitment, it gets extracted into a task system — it does not stay in the chat as its own record. A Slack thread is a conversation log, not a project-management tool, no matter how load-bearing it has quietly become for your team.
This matters because chat apps are built for retrieval by search, not by recall. Six weeks from now, you won't remember that the commitment lives in a thread titled "quick q" in a channel you've since muted. You'll remember there was a commitment, and then spend fifteen minutes searching for it — or worse, miss it entirely.
David Allen's Getting Things Done (GTD) methodology calls this "capturing" — the discipline of getting every open commitment out of your head (or, in this case, out of an ephemeral stream) and into a trusted system the instant it appears. Allen's core claim, adopted widely across knowledge-work teams over the past two decades, is that unresolved commitments create low-grade cognitive drag whether or not you're actively thinking about them.
The extraction itself should take under fifteen seconds:
- Spot the commitment — a deadline, an ask, a "can you," a promise you just made in a reply.
- Name it as a task, not a copy-paste of the message ("Send Q3 pricing draft to Maria" beats "re: pricing").
- Attach the deadline and the source so you can trace it back to the original thread if context matters later.
- Leave the message where it is. You don't need to pin it, star it, or reply "on it" as a memory aid — the task system is now the memory.
Where Prodinja fits into this rule
This is the one place a dedicated tool earns its keep, because sticky notes and starred messages don't survive a busy week. When a Slack message or email genuinely contains a commitment, Prodinja's Reminders are built to let you pull that commitment out of the stream immediately, so the original thread stops being the only place the promise is recorded. It's a small, honest piece of the workflow — not a replacement for the triage judgment above, just a place for what survives the triage to actually live.
A Batching Cadence You Can Sustain Without Missing Real Signal
Three fixed check-in windows a day — morning, post-lunch, late afternoon — outperform constant checking because each window comes with enough accumulated messages to triage efficiently, and the time between windows is protected for actual work. Continuous checking feels more responsive; it is measurably worse for both response quality and deep work.
Psychologist Gloria Mark's attention research at UC Irvine has tracked how people use computers for years, and her findings are blunt: once attention is interrupted, it typically takes on the order of 20-25 minutes to fully return to the original task. Every unbatched check of Slack is a small tax on whatever you were doing before you opened it — paid whether or not the message was actually urgent.
| Window | Duration | What happens |
|---|---|---|
| Morning (before first meeting) | 15-20 min | Triage overnight messages; extract any commitments; flag Tier 1 items needing same-day reply |
| Post-lunch | 10-15 min | Second pass on Tier 1 and 2; clear anything time-sensitive that arrived mid-morning |
| Late afternoon | 15-20 min | Final sweep; confirm nothing from Tier 1 is left hanging overnight |
Outside those windows, close the apps — not just mute them. A visible icon with a badge count, even silenced, still pulls attention every time your eyes pass over it. This is the same principle behind protecting deep work without missing real signal: the goal isn't zero interruption, it's interruption on your schedule instead of the sender's.
Set expectations so batching doesn't feel like going dark
Batching only works if your team knows to expect it — otherwise a three-hour gap in replies reads as unresponsive instead of intentional.
- Post your check-in windows in your Slack status or a pinned team-channel message.
- Tell direct reports and your manager explicitly which channel to use for anything genuinely urgent (a phone call or a specific @mention, not a general DM).
- Set an away message or status that states when you'll next check, not just that you're "busy."
Finding Real Signal Inside the Noise
Not every noisy channel is actually noise — sometimes a "low-priority" customer-support or sales channel is where the sharpest product signal is hiding, buried under volume that makes it look skippable. The fix isn't to read more of it; it's to apply a filter that tells you which messages in a high-volume channel are worth a second look.
A Jobs to Be Done lens is useful here: instead of skimming for keywords, skim for evidence of a job the customer is trying to get done and where the current product is failing them. A support message complaining about a specific button is noise; the same message read through a Jobs to Be Done lens might reveal a struggling moment worth escalating. The complaint is the surface; the job underneath it is the signal.
Mapping recurring complaints against a customer journey emotion curve does similar work at the pattern level. If the same friction keeps surfacing in Slack around the same stage of the journey, that's not one noisy message — it's a trend worth a proper look, which a structured customer journey exercise can surface far more reliably than a gut feeling formed from scrolling.
Key Takeaways
- Sort before you read. Every message is noise, signal, or a commitment — decide which, then act (or don't) accordingly, instead of reading top-to-bottom in arrival order.
- Tier your channels into must-see, scan-daily, and archive-only, and let notification settings enforce the tiers automatically instead of relying on willpower.
- Unread counts are not obligations. A high badge count measures arrival, not importance — treat "nothing important sat unactioned" as the real goal.
- Commitments leave the chat app immediately. A thread is a conversation log, not a task tracker; extract deadlines and promises into a trusted system within seconds of spotting them.
- Batch, don't graze. Three fixed check-in windows a day beat continuous checking on both response quality and protected time for real work.
- Some "noisy" channels hide real signal — apply a Jobs to Be Done or customer journey lens to high-volume customer channels instead of skimming past them.
- Re-tier quarterly. Channel importance shifts with your projects; a system set up once and never revisited quietly decays back into noise.
Frequently Asked Questions
How do I manage Slack as a product manager without missing important messages?
Tier every channel into must-see, scan-daily, and archive-only, then set notifications to match — full alerts only on Tier 1. Batch-check Tier 2 and 3 two to three times a day instead of continuously, and rely on the tiering, not vigilance, to catch what matters.
What's a good PM email system for a high-volume inbox?
Route email into three working folders — Action, Reference, and Watching — using filters so automated mail and CC-only threads never touch your primary inbox. Check Action first in each batch window, and stop treating "0 unread" as success; an untouched newsletter and an ignored decision request are not the same problem.
How often should a PM check Slack and email?
Three fixed windows a day — morning, post-lunch, and late afternoon — is a sustainable cadence for most PM roles, with Tier 1 channels getting real-time alerts in between. Constant checking feels responsive but costs roughly 20-25 minutes of refocus time per interruption, according to attention research from Gloria Mark at UC Irvine.
How do I make sure commitments made in Slack don't get lost?
The moment a message contains a deadline, a promise, or a "can you," extract it into a task, calendar block, or reminder system immediately — never leave it as a bookmark or starred message inside the chat. Tools built for this, including Prodinja's Reminders, exist specifically so the thread doesn't have to double as your memory.
Is it bad to leave Slack messages unread?
No — an unread badge is a queue, not a debt. Leaving Tier 3 (archive-only) channels unread indefinitely is the system working as intended; the only messages that need same-day attention are the ones that actually contain signal or a commitment aimed at you.
Building a durable version of this system is easier once it sits inside a broader PM productivity toolkit rather than as an isolated Slack hack — notification discipline, calendar defaults, and a tool audit all reinforce each other. Get the triage funnel right, and the unread badge stops being a source of anxiety and goes back to being what it always should have been: a number.