Shared inbox: which setup your team actually needs
support@ forwarding into one person's Gmail is a single point of failure. Here is every way to fix it, with the cost model of each and one honest recommendation per team size.
The cheapest way to run support@ is to forward it into one person's personal inbox, and it fails twice over: nobody else can see the queue, and the forward itself loses mail without telling you.
A shared inbox fixes that. It is one address (support@, sales@, hello@) that several people can read and answer without passing a password around. Every option in this article gets you there. The differences are price, clunkiness, and whether replies go out as the address or as the individual. This article assumes mail on your own domain already works; if it does not, start with our business email setup guide and come back.
Why support@ in one person's inbox fails
There are three standard workarounds, and each breaks in its own way.
Password sharing is the first. Two people logging into one mailbox defeats two-factor authentication, makes offboarding a fire drill (change the password, tell everyone, update every saved session), and leaves no record of who said what to a customer.
Forwarding is the second, and it fails quietly. We learned this in production: in June 2026, Gmail silently rejected 73% of the mail our own system auto-forwarded, because forwarding breaks DMARC alignment. Nothing errored on our side. The mail simply vanished. A support@ rule that forwards to someone's personal Gmail fails the same way, for some unpredictable fraction of messages, with no bounce to warn you.
The third workaround is doing nothing: support@ is one employee. That works until they take a vacation, and then your support queue takes one too.
Shared inbox for Gmail, option 1: Groups Collaborative Inbox
Google's real answer for teams is not in Gmail at all. It is Google Groups with the Collaborative Inbox feature switched on, included free with every Workspace plan. The group address does not consume a paid seat.
The setup, concretely:
- In the Admin console (or at groups.google.com on a Workspace account), create a group named
support@yourdomain.com. - In the group's settings, turn on "Collaborative Inbox" under additional Google Groups features.
- Set "Who can post" to "Anyone on the web," or customer mail will be rejected.
- Add your teammates as members with permission to moderate metadata, so they can assign conversations and mark them complete.
You get assignment, "mark as complete," and a shared history for $0. Now the clunk. The whole thing lives at groups.google.com in a forum-style interface, not in anyone's Gmail. Search is weaker than Gmail's. There are no shared drafts and no warning when two people answer the same message. Worst of all, replies default to the member's own address; answering as support@ requires each person to configure "Send mail as" in their own Gmail settings, one by one. Teams tolerate this at low volume. Nobody loves it.
Shared inbox for Gmail, option 2: delegation
Delegation grants other accounts access to one mailbox, reply rights included. Personal Gmail allows up to 10 delegates; Workspace accounts have allowed up to 1,000 since Google raised the limit in January 2021, with about 40 able to work concurrently under typical use (Google's published figures).
The catch sits in the From line: delegated replies are visibly marked as sent by the delegate on behalf of the owner. There is no assignment, no status, no queue. Delegation was built for an assistant managing an executive's mailbox, and for that it is the right tool. As a team support channel it is a hack.
Outlook and Microsoft 365 shared mailboxes
Microsoft's version is cleaner. A shared mailbox in Exchange Online needs no license of its own up to 50 GB of storage; past that it needs an Exchange Online Plan 2 license for the 100 GB tier (Microsoft's Exchange Online service description, checked August 2026). The real requirement hides one step back: every person who opens the shared mailbox must hold their own Exchange Online license, so this is only free inside an organization already paying for Microsoft 365.
Send As and Send on Behalf both work and are set by an admin, so replies can go out cleanly as support@. What you do not get is workflow: read and unread state is all the coordination there is, with no way to see who owns which conversation. For a Microsoft shop with modest volume, this is the best free option. The ownership problem stays yours to solve in chat.
Dedicated shared inbox software: Front and Hiver
When email becomes a support operation, the dedicated tools earn their keep: per-message assignment, collision detection (a banner when a colleague is already typing), SLA timers, CSAT surveys, and analytics.
They charge per seat, and that is the whole story of the bill. Neither one is in the 14-provider pricing dataset behind this site, so we will not quote their tiers from memory: open both pricing pages, and read the per-seat rate and the seat minimum before you budget.
Then do the multiplication, because it decides this. A five-person team on a $25 seat is $1,500 a year; the same team on a $65 seat is $3,900. Those are defensible numbers for a ten-plus-person support desk with contractual response times. Below that, you are paying help-desk prices for what is really a rota, and every hire adds a line to the invoice.
The LetterDuck model: every address is already shared
We built LetterDuck around a different unit. An address is a routing rule, not a seat. A catch-all routes every address on the domain into one threaded inbox, so support@, billing@, and any alias you invent next week are shared from the moment they exist. There is no provisioning step because there is nothing to provision.
Anyone in the workspace sees every thread. Replies go out as whichever address the conversation belongs to, with In-Reply-To and References headers set so the exchange threads correctly on the customer's side. The price is $50 per domain, flat, and hiring your fourth person changes nothing.
The honest limitation: LetterDuck has no per-message assignment, no collision detection, no SLA reporting. If you run a support desk that needs those, buy Front or Hiver and do not look back. LetterDuck is for teams of roughly one to ten who want every address on the domain answered from one place without per-seat pricing. It is the system we run our own mail on, and the inbound path behind it has been in production since June 2026 with a watchdog probe round-tripping it every two hours.
The decision table
| Team size and situation | Pick | Why |
|---|---|---|
| Solo founder | Aliases into one mailbox | Nothing to share yet; folders and aliases cover it |
| Executive plus assistant | Gmail delegation | Built for exactly this shape; the sent-by label is fine here |
| 2-3 people on Google Workspace, light volume | Groups Collaborative Inbox | $0 and workable if you accept the groups.google.com interface |
| 2-10 people already on Microsoft 365 | Outlook shared mailbox | Free under 50 GB; solve ownership in chat |
| 1-10 people, own domain, several addresses to cover | LetterDuck | $50 per domain flat; every address shared by default, reply as any of them. Waitlist only for now |
| 10+ people or contractual response times | Front or Hiver | Assignment, collision detection, SLAs, and analytics are worth per-seat money at this scale |
Where to start
Count two things: the people who need to answer mail, and the addresses they answer for. One person, any number of addresses: aliases. Many people, one executive mailbox: delegation. Many people, many addresses, under ten humans: pick flat pricing over seats, because the per-seat bill grows exactly when your margins are thinnest. And if the count crosses ten, or a contract mentions response times, pay for a real help desk without guilt. The expensive mistake is not choosing the wrong tier; it is leaving support@ pointed at one person's inbox for another quarter.