An omnichannel inbox puts email, WhatsApp and chat with the same customer in one place. Two products are sold under that name. One shows the channels side by side in one window, each with its own conversations and history. The other keeps one record per client, so a person who emailed on Tuesday and messaged on Thursday is one relationship. Only the second removes work.
Ask a vendor one question: if the same person emails and then messages, do I get one record or two? Side-by-side channels are common. One client record spanning Gmail, Outlook, IMAP and WhatsApp is rarer, and it is what stops a promise made on one channel going missing on the other. Test it in a trial with a real contact.
We build Sektra, which keeps one record per client across email and WhatsApp. The vendor facts below were checked on each vendor's site in October 2026.
What is an omnichannel inbox?
The term covers four levels, and vendors use the same words for all of them. Working out which level a demo is showing you is most of the evaluation.
- One login. Both channels are in one product, in separate sections. This is the version sold most often as unified, and the gain is one login.
- One queue. Conversations from every channel arrive in a single list the team can assign, tag and filter together. This is useful, because WhatsApp stops being missed for sitting in a different section.
- One contact. The same person is recognized across channels, so their email history and their messages sit on one timeline. This is where the tool starts doing work a person was doing.
- One record. On top of the timeline, the tool keeps what was promised and what is outstanding, whichever channel it happened on. Few products reach this level.
A request for one inbox is a request for the third or fourth level. The first two are what the category meant for a decade, so a demo that shows them is habit more than deception.
What is omnichannel messaging?
Omnichannel messaging is carrying one customer conversation across channels, so a question asked on WhatsApp can be answered by email and the customer experiences one thread. Multichannel means the business is reachable on several channels. Omnichannel means those channels share state. Customer communication management, the vendors' term for the whole system, adds the policies: who owns a conversation, how fast it is answered, and what is recorded.
For a client team the practical test is small. A client sends a document by WhatsApp and asks a question by email an hour later. Whoever answers the email should see the document without asking. If the tool makes that true, it is omnichannel in the sense that matters.
Why is WhatsApp the hard channel?
Email, whether Gmail, Outlook or an IMAP mailbox, is easy to add to an inbox, and so is web chat. WhatsApp is harder, for structural reasons.
- A business number is not a mailbox. To be shared by a team, a WhatsApp Business number has to be on the WhatsApp Business Platform, through Meta's Cloud API or a Business Solution Provider. The free WhatsApp Business app runs on one phone, with a few linked devices for extra screens. It gives a team no owner, notes or assignment, so the number moves onto the platform first.
- There is a 24-hour window. Once 24 hours have passed since the customer's last message, the business cannot send free-form text. It sends a pre-approved template, which Meta bills per delivered message. Follow-up works differently from email because of it.
- Identity is a phone number. Matching a WhatsApp number to the person whose email you already hold is not automatic. If the tool cannot do it, you get two records for one client, which is the problem you set out to solve.
- History does not arrive by default. Connecting a number does not usually bring the conversations that happened before. Ask about import.
The platform, the window and templates are covered in the WhatsApp shared inbox guide and on the WhatsApp integration page. For an evaluation, the short version is that a vendor who has solved WhatsApp will raise the 24-hour window unprompted.
Which vendors unify channels, and how far do they go?
Public capability as checked on each vendor's site in October 2026. Tiers change what is included, so check against the plan you would buy.
| Tool | Channels in one queue | Same person across channels |
|---|---|---|
| Sektra | Gmail, Outlook, IMAP and WhatsApp Business, synced two-way | One client record covering every channel, with commitments and open questions pulled from each |
| Front | Email, with WhatsApp Business on the Professional and Enterprise plans | Contact-level history within the app; no record of what was promised |
| Intercom | An omnichannel inbox around chat, email and phone, with WhatsApp through an integration | Contact profiles built around support tickets and the Fin AI agent |
| Missive | Email, SMS, WhatsApp, Instagram, Messenger and live chat | Conversations stay per channel, with shared team context |
| Help Scout | Email, SMS, social, chat and voice; WhatsApp is not named | Customer profile with conversation history, built for support |
| Hiver | Email in the Gmail product; WhatsApp listed for Hiver Omni, a separate omnichannel platform | Ticket-level history |
| Gmelius | Email, as a Gmail extension | No second channel to unify |
Two things stand out. Channel coverage no longer separates anyone, since WhatsApp is on or near every list. And nearly everything unifies at the conversation or ticket level, which suits a support queue. Account work needs the relationship level, and Sektra keeps one record per client across channels with the commitments and open questions in it.
What does a client switching channels break?
Test this part in a trial, because it is where side-by-side setups fail in ways a demo never shows.
- The commitment gets orphaned. You promise something by email on Monday. The client asks about it on WhatsApp on Friday. Whoever picks up the message has no reason to know a promise exists, and the client repeats themselves.
- Ownership splits. The account manager owns the email thread and whoever is on the messaging rota picks up the WhatsApp. Two people, one client, no shared view of what the other said.
- Reporting double counts. Two conversations with the same person about the same thing become two entries in whatever you report on, so response numbers look better than they are.
- Quiet goes unnoticed. A client who stops emailing but keeps messaging looks like a risk in one system and a healthy account in the other. Neither picture is true.
The last one is the expensive one and the slowest to show up. A relationship gone quiet on every channel at once is the signal that matters, and it can only be computed if the channels are one record.
How do you evaluate an omnichannel inbox in a one-week trial?
Five tests, all of which run with one real client and about an hour of attention.
- Email a contact from the tool, then message the same person on WhatsApp. Do you see one record or two? This one test is most of the evaluation.
- Make a commitment in the email, then ask the tool what you owe that client. If the answer covers only one channel, you have side-by-side.
- Have a colleague reply on the other channel. Does your view update, and can you tell what they said without switching sections?
- Wait 25 hours and try to send a WhatsApp message. If the tool does not surface the template requirement at that moment, your team will hit it at the worst time.
- Connect a mailbox with real history and check whether past conversations arrive. A tool briefed from day one is trusted in week one; the other kind is trusted in month three.
If a vendor cannot pass those five in a trial, that is the answer, and a cheaper one than month three. How to run a trial without disrupting the team is covered in switching without migrating. The account management page covers what the record is for.
When do you not need an omnichannel inbox?
Unification has a cost. If your clients only ever email you, adding a messaging channel because a tool supports it creates a new place to miss things. If inbound is mostly one-off questions from people you will never hear from again, you have a queue, and the ticket model matters more than the client record. If the team is three people in one room, the coordination problem is solved by talking.
The case for unifying is specific. You have named clients, they reach you on more than one channel, and a dropped promise costs retained revenue. If that is you, unify at the record level, because unifying at the window level moves the problem without solving it.
