A seller I watched last winter had four things open at once. Instagram on her phone. WhatsApp Web on the laptop. A Telegram channel she posted to twice a day. And the chat widget on her own site, pinging away in a browser tab she had muted weeks earlier.
A customer asked about a jacket size in Instagram DMs. Twenty minutes later the same person asked the same question on WhatsApp, because that was the number in her bio. She answered the WhatsApp one first, then went back and answered the Instagram one slightly differently. One said "the large runs a bit big." The other said "take the large, it's true to size." Small mismatch. The buyer noticed.
She didn't lose that sale. She lost the next three, from people who never got a second reply at all, because by then she was three tabs away.
That is the thing a multichannel inbox for sellers is supposed to fix. Most of the tools sold under that name fix half of it and quietly leave the expensive half alone.
A multichannel inbox is not four tabs in a nicer wrapper
Almost every tool in this category starts the same way: it pulls Instagram, WhatsApp, Telegram and your website chat into one screen, gives each conversation a little coloured badge, and calls it done. That is a real improvement over alt-tabbing. It is also the easy part.
The hard part is memory. If someone talked to you on Instagram on Sunday and messages your website widget on Wednesday, does the tool know it's the same person? Does whoever replies on Wednesday see Sunday's conversation without going to look for it? If the answer is no, you don't have a unified inbox. You have a screen with four inboxes stacked on it, and the mistakes just happen in a smaller area.
My rule of thumb: a multichannel inbox is worth paying for when the conversation history follows the customer, not the channel.
The tab tax nobody puts on the invoice
Say you get 60 conversations a day spread across four channels. Nothing dramatic, a normal week for a mid-sized Instagram shop with a site attached.
Each switch between apps costs you something small: reloading where you were, re-reading the last three messages, remembering whether you already sent the price. Call it 20 seconds. Sixty conversations, and realistically you touch most of them two or three times before they close. That is roughly 150 switches a day, and about 50 minutes of your working day spent doing nothing but re-orienting yourself.
Fifty minutes is annoying. The part that actually costs money is what happens to the messages you didn't get to during those fifty minutes. Buyers do not wait politely. We've written before about how sharply conversion falls once a reply drifts past the first few minutes, and the drop is steeper than most sellers expect.
So the tab tax gets paid twice. Once in your time, once in the sales that went cold while you were paying it.
Your customer doesn't know your channels are separate
This is the mental shift that makes everything else make sense. To you, Instagram and WhatsApp are two systems with two logins and two sets of rules. To a buyer, they're both "messaging the shop."
People move channels for boring reasons. Instagram DMs get buried under story replies. WhatsApp is where their thumb already is. They saw your Telegram post at midnight and replied there. They found your site through search and used the widget because it was in front of them.
When they switch, they expect you to remember. Nobody re-introduces themselves. They just say "so is it in stock?" and wait, and if the answer comes back as "hi! which product do you mean?" you have just told them, politely, that they're starting over.
Three questions decide most sales in messaging: price, availability, and how long delivery takes. When those answers arrive late or inconsistently, the deal dies quietly, which is exactly the pattern behind the unanswered questions that kill orders without anyone complaining.
Shared memory is the whole feature
Here's what "shared memory" looks like in practice, using that jacket customer.
She asks about sizing on Instagram. Two days later she opens WhatsApp and writes "is the black one back yet?" A tool with shared memory answers with the black jacket, in her size, at the price she was already quoted, and mentions that the sizing note from Tuesday still applies. A tool without it answers "which product?" and she goes to look at someone else's page.
Same message. Same shop. Completely different outcome, and the only difference is whether the history travelled with her.
The same logic applies to the assistant answering on your behalf. One assistant that reads all four channels and knows your catalogue, your prices and your policies will behave consistently. Four separate bots, each set up on a different afternoon with slightly different instructions, will drift apart within a month. You'll find out when a customer screenshots two answers side by side.
What to actually check before you buy one
Most demos look identical. These are the questions that separate them:
- Does history merge across channels? Ask them to show you one customer who wrote on two different platforms. If the demo can't do it live, assume it can't.
- Is there one set of product knowledge, or one per channel? You want to update a price in a single place and have every channel change at once.
- Can a human take over mid-conversation? And when they do, does the assistant stop talking, or keep replying over the top of your colleague?
- What happens on the channel you use least? Plenty of tools treat Instagram as the real product and everything else as a checkbox. Test the quiet channel, not the loud one.
- Does it work in your language, at your hours? Including the messages that arrive at 1 a.m., which is when a surprising share of buying decisions get made.
- Who owns the conversation data? If you leave in a year, do you take the history with you?
If you're weighing several options, it's worth reading through how to compare AI DM assistants without buying the wrong one before you sit through another demo.
Set it up in the order that hurts least
Don't connect everything on day one. That's how people end up switching it all off by Friday.
- Start with your busiest channel. For most sellers that's Instagram DMs. Get the answers right there first.
- Feed it the boring stuff. Sizes, prices, shipping times, return policy, payment methods. This is 70% of what anyone asks.
- Read the first hundred conversations yourself. Every wrong answer is a gap in what you gave it, and it takes about a week to close most of them.
- Add the second channel. WhatsApp or your website widget, whichever brings more real buyers.
- Then the rest. By now the knowledge is shared, so each new channel takes minutes rather than a weekend.
Adding channels one at a time also tells you something useful: which ones actually produce orders. Sellers are often wrong about this. The channel that feels busiest is not always the one paying for itself.
Keep a human door open
An assistant handling every routine question all day is a good thing. An assistant that refuses to let go when a customer gets frustrated is not.
Set a clear handoff rule: an angry customer, a complaint about a damaged order, a discount request outside your rules, anyone who types "human" in any spelling. Those go to you, with the whole conversation attached, on whatever channel the customer is already using. You should never have to open a different app to answer a handoff.
That last detail is what separates a real unified inbox from a demo. If escalation means "now go to Instagram and find her," nothing was unified.
Where this leaves you
The pitch for a multichannel inbox is usually about tidiness. Fewer apps, one screen, a calmer morning. Fine, but that's not where the money is.
The money is in consistency. Same answer on every channel, at 2 p.m. and at 2 a.m., whether it came from you or from your assistant, whether the customer wrote once or five times across three platforms. That is what makes a small shop feel like a business worth sending money to.
If you want that without stitching four tools together yourself, Vardast connects Instagram, WhatsApp, Telegram and your website to one assistant that shares a single memory and one set of product knowledge, answers in your voice, and hands the conversation to you the moment a person needs to step in.
Start with your busiest channel this week. Read what it gets wrong. Then add the next one. That's the whole method.