Question 1

Which marketplaces and channels does ReplyIQ support?

Accepted Answer

On launch ReplyIQ is first-class on Shopify, Amazon Seller Central, eBay, and email — the four inboxes a multi-channel seller is most likely running from a single laptop. Growth tier adds Chat and SMS onto the same inbox; Scale adds Social and Voice. Each channel keeps its own messaging rules (no off-platform links on Amazon, case-aware replies on eBay, return windows on Shopify, threaded replies on email) and ReplyIQ stays inside them.

Question 2

How does data move between my channels and ReplyIQ?

Accepted Answer

Inbound messages are pulled through the channel-native sync — webhooks for Shopify, the Buyer-Seller Messaging API for Amazon, eBay Messages for eBay, and IMAP polling plus push for email. Each message is joined to its order, listing, return case, or refund record, attached to the conversation thread, and stored against the originating customer profile. From there it sits in your unified inbox and is replied to or escalated on-policy — the full loop without copy-paste between tabs. See /integrations for the per-channel walkthrough.

Question 3

When does the AI escalate a conversation instead of replying?

Accepted Answer

Escalations fire on hard, auditable triggers: a refund amount above the configured tier cap, legally-loaded language, a chargeback mention, a sentiment crash past a defined threshold, or any custom keyword set your team defines. Every escalation carries its reason as a note attached to the ticket, so your agents know exactly why a conversation was handed off — never a vague "this looked risky".

Question 4

Where do escalations go?

Accepted Answer

Each escalation rule has an explicit destination — an escalation queue, an email alias, a Slack channel, or a named agent — configured in merchant settings. Refund escalations can land in one queue, legal-language escalations in another, and chargebacks in a third, so the right person is paged without anyone triaging the inbound stream first. Destinations are per-rule and per-store.

Question 5

What changes after we go live?

Accepted Answer

On your side, almost nothing. ReplyIQ keeps reading your live merchant policies, so when your team edits a return window, a refund cap, or a tone-of-voice guideline in the policy store, the AI starts honouring it on the very next reply — no deployment, no model retraining, no migration. You also get a weekly digest covering resolved tickets, escalations, and policy coverage gaps so you can tune over time without exporting CSVs.

Question 6

Where does my customer data live, and how is it protected?

Accepted Answer

EU customer data is stored in our Frankfurt Postgres cluster; US customer data stays in Virginia. Conversation text, ticket metadata, and policy documents are in-region by default; cross-region replication is opt-in and only used for failover — never for analytics. PII is encrypted at rest with per-tenant KMS keys, and single-tenant data export plus full deletion are available on request through the API or your account team.

Question 7

How does the AI learn my store's policies?

Accepted Answer

ReplyIQ reads from your merchant policy library — return windows, refund tiers, shipping cutoffs, brand voice, do / do-not lists — the same way a trained agent would, before every reply. The policies themselves ARE the knowledge base — there is no parallel model to retrain on the side. Edit a return window in the policy store and the AI picks it up on the next message; there is no lag and no black box.

Question 8

How are per-store cancellation and refund rules respected?

Accepted Answer

Refund tier overrides are the standard mechanism. By default Tier 1 (small, low-risk) returns are settled by the AI within the configured cap; Tier 2 and Tier 3 refunds — high-value items, edge cases, or anything your team flags explicitly — route to your agents for approval with the full conversation context pre-filled. Every store can edit its own tier caps, escalation destinations, and approval requirements from merchant settings without contacting us, and overrides apply per channel so Amazon can run a tighter cap than Shopify if your risk profile differs between marketplaces.

Question 9

How is ReplyIQ different from a generic email tool?

Accepted Answer

Generic email tools bolt AI onto a single inbox and call that multi-channel. ReplyIQ is multi-channel-native — the same conversation thread runs across Shopify, Amazon Seller Central, eBay, email, Chat, SMS, Social, and Voice from day one, with each channel keeping its native rules. The policy file IS the knowledge base: edit a return window, refund cap, or tone-of-voice guideline and the very next reply honours it — no parallel model to retrain, no deployment in between. Hard escalation triggers (refund cap, legal language, chargeback, sentiment crash) hand the conversation to a named queue with the reason attached, instead of letting the AI guess.

Question 10

How long is conversation data retained, and what is our GDPR posture?

Accepted Answer

Conversation text, ticket metadata, and policy documents are stored in-region by default — EU customers on Frankfurt, US customers on Virginia — and cross-region replication is opt-in, used only for failover. PII is encrypted at rest with per-tenant KMS keys, and every tenant gets a single-tenant export of their full dataset plus a full-deletion endpoint on request, through the API or via their account team. The same export endpoint serves GDPR data-subject access requests, so fulfilling a DSAR does not require a separate tooling path.

Question 11

What doesn't ReplyIQ replace?

Accepted Answer

ReplyIQ sits beside your existing stack, not in front of it. The marketplace seller consoles stay the source of truth for fulfilment, inventory, and shipping — we read them, we do not replace the order pipeline. Shipping and catalog tooling (rate shoppers, OMS, PIM) keep owning their domain; ReplyIQ only consumes the order data they expose. Voice and SMS providers stay yours — we connect to the channels you already pay for rather than bundling a separate telecom stack. Custom CRM exports, accounting integrations, and BI pipelines continue to run on the data your stack already produces; ReplyIQ adds a reply-and-escalation layer on top.

Questions about ReplyIQ

Positioning

Data handling

Integration model

Still have questions? Start onboarding.