One Inbox for WhatsApp and Instagram: What to Check Before Several People Share a Number

Once more than two people answer the same number, the question stops being "can it send messages". It becomes "who is supposed to answer this one". Faisal Town, a major real-estate development in Islamabad, gets property enquiries constantly and they do not arrive sorted — plot availability, transfers, dues, construction, after-sales. The thing that actually changed for them was not automation. It was that ownership of a conversation became visible instead of assumed.

If several people answer the same WhatsApp and Instagram number, the thing you are shopping for is a shared team inbox: one queue, with assignment, ownership and history that outlive whoever happens to be on shift. Meta gives you a real version of this for free, and a lot of teams should just use it. This page covers what the free version does, five things worth testing before you pay for anything else, and the point where paying starts to make sense.

The three shapes this comes in

Vendors describe all three as "a shared inbox", which is how teams end up buying the wrong one.

Start here: what Business Suite already does free

This gets skipped by almost everyone selling you something, so it is worth being specific. Meta's own Inbox reads messages and comments from your Facebook Page, Messenger, Instagram and WhatsApp Business account together, and it is free. (Manage messages with Inbox)

• Assign a conversation to a specific teammate on desktop, if you are an admin, editor or moderator — and unassign from the same menu. (Assign conversations)
• Labels, notes, Move to Done and Mark as Follow Up for basic queue hygiene.
• Saved replies your team inserts by hand, and ice breakers shown before someone writes.
• Automated responses — instant reply, away message, and up to 5 custom keywords. (Automated responses)

Test 1 — Can you tell who owns a conversation?

Not who replied last. Who is *responsible* for it now. Open a chat someone else is handling and ask whether the tool shows you an owner, or whether you have to read the thread and guess. Then have that person go on leave and see what happens to their conversations. This is the single most common reason a growing team outgrows a setup that was working fine.

Test 2 — Can a manager see what happened without asking?

Ask for the thing a manager actually wants: which enquiries came in yesterday, who took them, which are still open, and which got no reply at all. If answering that means someone screenshotting their phone or exporting a CSV, the tool has an inbox but not a system. At Faisal Town this was the real unlock — management could see team performance in the same place the work happened, rather than requesting it.

Test 3 — Can a reply depend on the previous answer?

This is the hard line between the free tools and a platform, and it is narrower than most marketing implies. A Business Suite automation is one trigger producing one response. It cannot ask "are you enquiring about a new plot or an existing one?", read the answer, and route accordingly. If your replies are answers, the free tools answer them. If your replies need to make decisions, no number of keyword rules gets you there.

Test it by writing down one real conversation that needs two questions before it can be routed, then try to build it.

Test 4 — When does the AI stop and a human start?

Any vendor will demo AI answering a question. Ask the less flattering one: what happens when it does not know, or when the customer gets angry? You want a rule you control — specific words, sentiment, value, time of day — not a bot that keeps talking because handing over was never designed. Ask to see the handover configured, not described.

Test 5 — Can you tell which conversations became revenue?

Ask this one bluntly, because the honest answer from most of the industry — ours included — is "partly". Delivery, read and reply analytics are standard. Tying a specific conversation to a closed sale usually is not, because the sale closes in your CRM or your POS, not in the chat.

What to settle before you sign: whether labels or conversation states can act as pipeline stages, and whether the platform can push those into the system where revenue is actually recorded. A vendor claiming clean revenue attribution out of the box, with no integration work, is overselling.

The limits that bind every tool on this list

Some ceilings get blamed on whichever inbox you are using, but they are Meta's and they apply to all of them — free tools and paid platforms alike. Worth knowing before you switch and find the problem followed you.

• 24 hours to reply. Once a customer messages you, you have a 24-hour window to respond, and promotional content is allowed inside it. (Messaging policy)
• 7 days with a human. The human agent tag extends that window, strictly so a person can resolve what the standard window could not. It is not a way to keep automation running. (Human Agent)
• Keyword matching is literal. In Business Suite, custom keywords are case-sensitive and must match exactly, so a rule written as "PRICE" never answers someone who types "price".

Where WABPO fits, and where it does not

We are built for the row at the bottom of that first table: WhatsApp, Instagram and Messenger in one queue with assignment and ownership, a visual builder where replies branch on answers, an AI knowledge base with a handover rule you set, and integrations that put order and booking data into the thread. Meta bills your own account at rate-card prices with no markup from us.

Where we are the wrong answer: if your customers live on Telegram, LINE or channels outside Meta, respond.io has channel breadth we do not match, along with more mature enterprise workspace controls. If your main job is turning Reel comments into DMs, ManyChat does that and we do not. And if Business Suite covers the five tests above, it is free and it is from Meta — use it.