A customer sends an email, calls two days later because nobody got back to them, then sends a message on the messenger they had open anyway. One request has become three. Inside the company they sit in three different systems, handled by two different people, neither of whom knows the other two exist.
The channel is the customer's choice, not yours
The common response is to close channels. No messages by messenger, enquiries through the form only. That works in theory and fails in practice: people who have a number will call, and people who see an open channel will use it. A closed channel does not become empty, it becomes unwatched.
The more durable answer turns it around. The customer picks the channel, the company brings them together. What matters is not which route something arrives by, but whether it lands on the same case.
What “one inbox” actually means
A shared inbox is not merely a list with everything stacked underneath each other. Three properties make the difference.
Attached to the customer. The email, the call and the messenger message hang on the same customer record. Whoever opens the conversation sees the history without searching.
Visible ownership. It is clear who is handling it. Without that you get the second classic failure after things being missed: two contradictory answers to the same question.
One history instead of many. The phone call leaves a trace too, with a time and a note. Otherwise the channel where the important things get agreed stays the only one without a memory.
Where telephony makes the biggest difference
Telephony is the channel left out of digitisation longest, even though it is where binding commitments are made. When an incoming call identifies the customer and shows their open cases before the handset is picked up, the conversation changes. And when a note stays on the case afterwards, the commitment does not vanish when the call ends.
The second point concerns calls that reach nobody: outside business hours, over lunch, during a disruption. A voice assistant that takes the call, records the request and opens a case does not replace a proper conversation. It makes sure a missed call does not become a lost customer, which is a different ambition from replacing a switchboard.
The mistake worth avoiding
A consolidated inbox tempts you to automate everything that can be automated. Experience argues against it. Automatic replies are good where the question really is always the same: opening hours, delivery status, appointment confirmation. As soon as judgement is involved, an automatic reply costs more trust than it saves time.
The sensible split is therefore not “machine answers, human checks” but “machine sorts and prepares, human answers”. The time is saved in the sorting anyway, not in the wording.
Frequently asked questions
Do we have to change our phone number?
Usually not. Porting existing numbers is the normal case, but it depends on the losing provider and the existing contract, so it belongs settled early. Changing the number would be the most expensive part of the move for existing customers.
What about personal mailboxes?
They stay. What gets consolidated is the shared addresses: info, sales, support. Personal correspondence does not belong in a team inbox.
Does every member of staff see everything?
No, the same access rights apply as everywhere else in the system. A shared inbox is not a suspension of permissions, it is a different view onto them.
The AI telephony use case shows how this works with calls. The modules page gives the overview, and you can begin free of charge.