Most remote teams do not have a tools problem. They have a too many tools problem — a decision in Slack, a spec in Notion, a ticket in Linear, a recording in Zoom, and nobody able to reconstruct why anything was decided.
The useful question is not which apps are best. It is which four jobs a remote team has to cover, and what happens when one tool tries to do two of them.
The four jobs, and nothing else
| Job | What it must do | Failure symptom |
|---|---|---|
| Talk | Fast, disposable conversation | Decisions buried in scrollback |
| Decide | Durable record of what and why | Same argument every quarter |
| Track | Who owns what, by when | Work discovered late, or twice |
| See | Face-to-face when text fails | Long threads that should have been a call |
One tool per job. Two tools for one job is where teams lose things, because half the team uses one and half uses the other.
Talk: chat, and the rule that saves it
Slack, Teams and Discord all do this well enough that the choice rarely matters. Teams wins if you already pay for Microsoft 365. Slack wins on integrations and search. Discord wins on cost and on voice channels that people drop in and out of.
What matters far more than the choice is one rule: chat is not a record. The moment something is decided, it moves to the decide tool. Without that rule, your team’s institutional memory is a search box, and search boxes do not surface the message where someone said “fine, let’s do B” three months ago.
The practical version: a bot or a person posts the decision into the doc and links it back. Thirty seconds, and it removes most of the confusion remote teams complain about.
Decide: the tool most teams under-invest in
This is the docs layer — Notion, Confluence, Google Docs, or a repo of markdown files. Whatever you pick, it needs three properties: searchable, linkable, and edited in place rather than re-sent as attachments.
| Tool | Cost per user | Strength | Weakness |
|---|---|---|---|
| Notion | ~$10/mo | Docs and databases together | Slow at scale; weak permissions on lower tiers |
| Confluence | ~$6/mo | Real permissions and hierarchy | Heavy; nobody enjoys writing in it |
| Google Docs | ~$6/mo (Workspace) | Best editing and review experience | Poor structure — a folder is not a wiki |
| Markdown in Git | Free | Version history, reviewable changes | Only works if the whole team writes code |
The common failure is choosing Docs for the writing experience and ending up with 400 documents nobody can navigate. If you go that route, impose structure early — an index page that is genuinely maintained, not a Drive folder.
Track: pick by how your work arrives
- Work arrives as tickets from outside (support, clients, bugs) — Jira or Linear. You need queues, priorities and SLAs.
- Work arrives as projects you plan — Asana or ClickUp. You need timelines and dependencies more than queues.
- Work arrives as a short list you clear — a shared board in Trello, or honestly a shared task list. Do not install Jira for a team of four.
The mistake is adopting the tool used at someone’s last job, at a company ten times the size. Process tools cost time on every single task; the cost is only worth it when coordination is genuinely hard.
See: fewer meetings, better recorded
Zoom, Meet and Teams are interchangeable for most purposes. What differentiates a remote team is not the video tool but two habits around it.
Async video first. A three-minute recorded walkthrough replaces a large share of status meetings and lets people watch at 1.5× at a time that suits them. Loom popularised this; every major tool now has a version.
Every call produces a written artefact. Transcription plus a summary of decisions and owners, pasted into the decide tool. This is the one place where AI tooling earns its subscription outright — it is fast, and mistakes are obvious on reading.
A stack that works for most small teams
| Job | Under 10 people | 10–50 people |
|---|---|---|
| Talk | Slack free tier | Slack or Teams, paid |
| Decide | Notion | Notion or Confluence |
| Track | Linear or a Notion database | Linear or Jira |
| See | Meet + a transcription tool | Zoom + transcription |
If you are running the decide layer in Notion, its databases can also cover tracking for a small team — one tool, two jobs, and in this specific case it works because they share the same content. Relations are what make that possible; we explained them in Notion relations and rollups.
Common questions
How many tools is too many?
The test is not a number. It is whether a new joiner can be told, in one sentence per tool, what goes where. If explaining the stack takes more than four sentences, work is being lost between them.
Do we need a separate wiki and doc tool?
No. That is the classic two-tools-one-job trap. Pick one, and accept that it will be slightly worse at one of the two uses than a dedicated app would be.
What about time zones?
Time zones are a writing problem, not a tooling problem. Teams that write things down properly handle eight-hour gaps comfortably; teams that rely on being online at the same time struggle with two.
Is free tier enough for a small team?
Usually yes for talk and see. The decide layer is where free tiers hurt, because message and version history limits quietly delete your record of why things were decided — which is the entire point of that layer.
Leave a Reply