Internal Communication Tools: Reaching People Who Miss the Thread

The trigger for looking up internal communication tools is almost never that nobody is talking. It is that somebody missed something they needed, and the message was definitely sent. It was in the channel. Two people replied to it. The person who needed it was on leave, or on a different project that week, or reading a channel with forty other messages in it that day.

Adding a tool is the reflex response, and sometimes it is the right one. But a new tool only helps if the thing that failed was a missing channel. If the thing that failed was that an important message had the same weight as a lunch order, another app makes the problem worse by adding a place to miss things. Before choosing, it is worth being precise about which of five distinct jobs is currently unowned.

Five jobs, not one category

Product roundups list tools together because they all involve people communicating. They are not substitutes for each other. Each does one of five jobs, and each job fails in its own way.

Conversation is talking, quickly, with the option of not writing carefully. Chat platforms own this and own it well.

Broadcast is one message that must reach everyone, where the sender needs to know it arrived. Newsletters, intranets, announcement channels and employee apps live here.

Reference is the knowledge that should be written once and consulted many times. Handbooks, runbooks, policies, onboarding material.

Work state is who is doing what, how far along it is, and what is blocking it. Boards, trackers and schedules.

Record is what was decided and why, attached to the thing it was about so it can be found later by somebody who was not there.

Most teams have conversation covered several times over and record covered nowhere. That mismatch produces the exact failure that starts the search. A decision made in conversation stays in conversation, where it is findable only by people who know it happened and roughly when.

How the categories actually compare

Category	Job it does	Reaches	Findable later by	Typical failure
Chat platform	Conversation	People currently reading	Someone who knows what to search	Everything has equal weight, so important things scroll away
Intranet or employee app	Broadcast	Everyone, including non-desk staff	Someone browsing a structure	Nobody visits unless pushed there
Wiki or docs tool	Reference	People who go looking	Anyone, if it is maintained	Goes stale quietly, and stale is worse than absent
Email	Broadcast, externally too	Everyone with an address	The recipient only	Threads fork and the canonical version is lost
Board or tracker	Work state and record	People who open the work	Anyone who opens that work	Only holds what people put on it
Video and meeting tools	Conversation, synchronous	Whoever attends	Nobody, unless recorded and indexed	The output exists only in attendees' memories

Read down the "findable later" column. That is the column the original problem lives in. A tool that reaches everyone today and is unfindable next month solves broadcast and nothing else.

The non-desk question changes the answer

Teams where a meaningful share of people do not sit at a computer all day have a different shortlist. Chat platforms assume a notification gets read within the hour. For shift-based work, a message sent at 2pm may be seen at the start of the next shift, or not at all if history has rolled past. This is the case where a dedicated broadcast layer earns its place, and where anything important also has to exist in a place tied to the shift rather than to a moment in a stream.

Broadcast fails for a different reason than conversation

Conversation tools fail by burying things. Broadcast fails by not being read at all, and the two need different fixes.

The instinct is to make announcements louder: pin them, bold them, tag everyone. That works for the first few and stops working quickly, because the cost of an announcement is borne by every reader while the benefit accrues to the sender. Once a channel has trained people that most of what is marked important is not, nothing marked important gets read. This is the mechanism behind the common complaint that staff ignore announcements, and it is a consequence of how the channel has been used rather than a fact about the staff.

Two things fix it, and neither is a tool. The first is a hard rule about what goes out as a broadcast, with everything else demoted to a place people check when they want it. Four or five genuine announcements a month are read. Twenty are not. The second is that anything requiring action must also exist somewhere that keeps state, so an unread announcement is not the only record that the action was requested. A broadcast that says a form is due next Friday, with no corresponding item on anyone's board, depends entirely on a message being read on the day it arrived.

Acknowledgement is worth less than it looks

Tools that offer read receipts and required acknowledgement sell the reassurance that a message landed. What they actually record is that a button was pressed. For anything genuinely important, the thing worth tracking is the action, not the acknowledgement. Whether the form was submitted is visible without asking. Whether the announcement was read is not, and chasing it consumes the time of whoever is coordinating.

Free tiers decide how long your record lasts

The limits on free plans are not just a pricing question. They set the expiry date on anything written in that tool, which matters because most teams write their decisions in the cheapest tool they already have. All figures below were checked on the vendors' own pricing pages on 26 September 2026 and are worth rechecking before committing, because plans change.

Slack's free plan keeps 90 days of message history and allows up to ten app integrations. Unlimited history starts with the Pro plan. Anything written in chat on a free workspace stops existing as a record after roughly a quarter, which is longer than people remember and shorter than the life of most decisions.

Notion's free plan caps individual file uploads at 5MB and keeps page history for seven days, with the collaborative block limit lifting on the paid Plus plan. For reference material this is workable, since reference documents are rarely large and rarely need version history.

Microsoft Teams Essentials is listed at $4.00 per user per month paid yearly, with 10GB of cloud storage per user, and Microsoft 365 Business Basic at $7.00 per user per month paid yearly with 1TB per user. Teams tends to arrive bundled with a subscription an organisation already pays for, which is why it often wins on cost without being chosen on merit.

The pattern across all of them is that free tiers are generous on the conversation job and thin on the record job. Retention limits, version history and storage are the first things restricted, and those three are exactly what a record needs.

The test before adding another tool

There is one question that predicts whether a new tool will be used in three months. Does it remove a place, or add one?

A tool that removes a place is one where something that currently lives in two systems now lives in one. Fewer lookups, fewer things to check, less chance of missing something. A tool that adds a place means one more thing to read and one more place a message can hide. Adoption of the second kind decays quietly. People try it for a fortnight and drift back to whatever is already open on their screen.

This is why the honest first move is often consolidation rather than acquisition. If decisions are currently in chat, meeting notes, and a doc that somebody maintains, the fix is picking one of those three, not adding a fourth. The integrations question is part of the same calculation: a tool that can pull from what the team already uses removes a place, and one that requires parallel entry adds one.

Counting the cost people do not put in the spreadsheet

Per-seat pricing is the visible cost and usually the smaller one. The larger cost is the time spent checking multiple places, plus the cost of the things that get missed. Neither appears on an invoice, which is why the cheapest option on paper frequently loses.

One measurable proxy: count how many places a person has to check before they can answer "what is the current state of this project". If that number is three or more, the tool budget is not the problem.

There is a second cost that shows up later, and it belongs in the decision rather than in a review a year afterwards. Whatever holds the record becomes the hardest tool to leave, because leaving it means either exporting years of context into a shape another tool can read, or accepting that the history stops there. Checking the export path before adopting takes minutes and is the single cheapest piece of diligence available. Two specific things are worth confirming: whether the data comes out in a format something else can import, and whether attachments come with it or only as links that break once the account closes. A tool that is pleasant to use and impossible to leave is a more expensive commitment than the per-seat price suggests.

Where the record should actually live

The durable answer for the record job is to attach it to the work rather than to a date or a channel. A decision about a feature belongs on the feature. A change to a deadline belongs on the thing with the deadline. Found this way, it needs no search, because anyone who opens the work sees it in the state it was left.

That is also why boards and trackers quietly do the record job better than tools built for communication. The unit of organisation is the work, not the timestamp. When conversation sits next to the work instead of in a separate stream, the record forms as a by-product of talking, rather than requiring somebody to write a summary afterwards. Tools differ a lot in whether they support this: some keep comments on cards, some keep a separate chat, and some do both. The tool comparisons set out which of a team's current tools each option would actually replace, which is the question that matters more than any feature list.

What to change first

Pick the last thing that somebody genuinely missed, and work out which of the five jobs failed. Then move that one job into a single place and say so out loud, rather than adding a tool for it. If decisions are currently scattered across chat and meeting notes, Pinateca keeps chat next to the work and is free for up to five people and ten boards, which is enough to test whether the record job is the one that was unowned.

Q1. What is the difference between internal communication tools and collaboration tools?

The labels overlap and vendors use them interchangeably. The useful distinction is direction: communication tools move a message from a sender to readers, while collaboration tools hold a shared object that several people change. A chat platform and a newsletter are the first kind. A board, a document and a tracker are the second. Teams usually have too much of the first and not enough of the second.

Q2. How many internal communication tools should a small team have?

Enough to cover the jobs that are actually failing, which for most teams under twenty people is two or three. One for conversation, one that holds the work and the record attached to it, and reference material somewhere stable. Adding a fourth is worth doing only when it lets something else be switched off.

Q3. Is email still a reasonable internal communication tool?

For broadcast to everyone, including people without a company account, it remains the most reliable delivery mechanism. For anything that needs to be found again later it is the weakest option, because threads fork and each recipient ends up with a slightly different version of the record. The practical split is to announce by email and keep the canonical version somewhere with a single address.

Q4. Why do internal communication tools stop being used after a few months?

Usually because they added a place rather than removing one. If reporting can happen in either chat or the new tool, everyone chooses whichever is already open, and the new tool empties out. The other common cause is that people write into it and nothing comes back: no questions stop arriving, no meetings get shorter, so the writing stops.

Q5. How do you reach staff who do not use a computer at work?

Treat it as the broadcast job and solve it separately from conversation. That normally means a mobile app or a physical posting point, plus a rule that anything essential is also attached to the shift or task itself rather than only sent as a message. Assuming a notification will be read within the hour is the mistake that causes most missed information in shift-based teams.