It is a Tuesday afternoon. A minute ago your online store was taking orders and now it shows nothing but an error page. The platform says it is investigating, and all you can do is wait. Meanwhile the customers who were about to buy do not write to tell you: they close the tab and get on with their day. You will never know how many they were, which products they were looking for, or how much money went with them, because the outage took away the one place where they could discover and choose your products.
What if that error page were not the end of the journey? Instead of a plain "come back later", the shopper could reach a conversational assistant and carry on exploring the catalogue, get their questions answered, and build a wishlist of the products they want. That list can be saved, shared over WhatsApp or email, or picked back up when the store is available again to complete the purchase. The sale may be delayed by minutes or hours, but the intent to buy does not go down with the store.
The same conversational storefront, on a phone and on desktop — one agent, every screen.
It is worth being precise about this, because "the store is down" hides three separate losses that behave very differently.
The first loss is the one everybody thinks about and the one you can do least about. The other two are the ones that quietly cost more, and they are not actually caused by the outage. They are caused by putting every one of those functions behind the same front door.
On an online store, discovering products, getting questions answered and paying all happen in the same place. That feels like the normal way round — until that place stops working. Then you find out you had concentrated the entire customer journey into a single point of failure. It makes no difference whether someone only wanted to check that the blue trainer was in stock in a 42 or was one click from paying: both end up looking at the same error page.
A conversational storefront breaks that dependency. Instead of reading the catalogue straight from the ecommerce platform, it works from a synced copy that it keeps up to date. Even while the store is temporarily out of service, the customer can carry on discovering products, comparing them, getting answers and saving a wishlist to pick up when the checkout is available again. The purchase is interrupted; the conversation — and the intent to buy — is not.
Before
With a conversational storefront
That is the whole trick, and it is deliberately unglamorous. Not a backup shop. Not a second checkout. Just the observation that the questions which decide a sale do not need the shop to be online, and there is no good reason to let them die with it.
Nobody buys on AllHub. The payment, the order and the delivery all happen on your own store, on your own platform, under your own terms — that is true on a good day and it does not change on a bad one. So when your platform is down, the order cannot be placed. Not by us, not by anyone.
Any tool promising to "keep selling through an outage" in the transactional sense is either running its own checkout, which is a very different product with very different consequences, or it is overselling. We would rather say the narrower true thing: what survives is the conversation, the catalogue and the buyer's decision. The transaction waits.
Which turns out to be most of the value anyway. A shopper who has had their questions answered and picked their item is a sale that has already been made in every sense except the payment. A shopper who hit an error page at the moment of interest is nothing at all, and you will never even know they were there.
The worst outcome is not the outage. It is sending someone who is ready to buy into a broken tab with no explanation, because that is the version they remember. So the assistant checks, and when your storefront is not answering it says so plainly, once, at the top of the conversation:
Worth stating plainly: the conversational storefront does not replace your online store during the incident. It is not a second shop window and not a parallel checkout. What it keeps alive is product discovery — browsing the catalogue, comparing, asking and deciding — while the payment waits for your store to come back.
This step is yours. The store owner — or whoever looks after the store — has to customise the error page, or set up a redirect, so it says the store is temporarily out of service and offers a link to the conversational storefront. It matters that the customer understands why they arrived in that channel: someone dropped into a chat with no explanation feels lost and leaves, and the link gives back almost nothing of what it could.
Example — what the shopper reads
We are having a temporary problem with the store.
While it comes back, you can carry on exploring the catalogue and save the products you are interested in.
When the store is working again, your selection will be ready for you to complete the purchase.
With that explanation in front of them the rest takes care of itself: from the conversational storefront the customer carries on exploring the catalogue, gets their questions answered, and creates or shares a wishlist to complete the purchase once the store is available again. That is the entire objective — that the outage does not take the intent to buy down with it.
And one thing it does not say: that you will let them know when the store is back. The wishlist lives in the shopper's own browser, which is a deliberate privacy choice — no account, no email captured to look at a lamp. That also means there is genuinely nobody for us to write to, so the assistant does not pretend otherwise. It tells them their choice is saved and waiting here, which is the part that is true.
The check runs on a schedule and again the moment a shopper reaches for the buy button, so an outage that starts mid-session is caught in seconds rather than minutes. It is deliberately reluctant to declare your store down — one failed request is not an outage, and greying out the buy button on a shop that is selling perfectly well would be a far worse bug than the one it is preventing. Recovery is the opposite: the first successful check brings everything straight back.
The conversational storefront connects to your catalogue on Shopify, WooCommerce, Wix or Amazon and keeps its own copy of it. That copy is what the ai shopping assistant answers from, which is why it is still standing when the source is not. It is also why answers arrive in the time it takes to read them rather than the time it takes to query a busy platform.
Worth saying plainly: this is not a reason to choose your platform badly, and it is not insurance. Outages are rare and mostly short. The argument is simply that when one happens, the difference between a bad afternoon and a lost one is whether anybody was still able to answer your customers.
AllHub is in soft launch. The first ten stores on each platform get three months free. The price is not money — it is ten minutes on a call, once a week.
We are not collecting ten logos for a landing page. We are looking for stores with real buyers asking real questions, run by someone willing to say plainly when something does not work. What you get back is the founders' direct attention and the setup done with you rather than handed to you.
You cannot stop your platform from having a bad afternoon, and you should be suspicious of anyone selling you that. What you can decide is what happens to the person who was ready to buy during it. Right now the answer for most stores is an error page and a permanent silence. It does not have to be.
When your store goes down, your products do not have to. If yours has real buyers asking real questions, request early access — ten stores per platform, three months free, ten minutes of honesty a week.
Written by AllHub Team · AI Agents for Ecommerce
We build the AI agent team that sells, supports and grows ecommerce stores — EU-hosted, GDPR-first.
This article was created with the help of AI and reviewed by our team. We take great care over every post and every translation, but the odd mistake can still slip through. If you find one, write to us: you will be helping us improve.
AI-generated contentA model can fill a missing fact with a plausible answer. AllHub reduces that risk by separating verified commerce data from generated language.
AI-generated contentYes. A comparison can use correct facts and still reach a harmful conclusion when it ignores audience, positioning or the reason behind each difference.
AI-generated contentA safe agent does not replace your catalogue or decide what you sell. It uses your products, data and rules to help buyers understand them.
