Help Scout Alternative for Shopify Stores: Who Outgrows the Shared Inbox
Help Scout AI answers from your help docs, not your live orders, so order status still hits a human. A Help Scout alternative built for Shopify stores.
A customer emails at 11pm asking where order #1042 is. The tracking number is sitting right there in the Help Scout sidebar, three inches from the reply box, pulled straight out of Shopify. Nobody sees it until 8:40 the next morning, when someone opens the inbox, glances right, copies the number, and sends a two-line reply.
That reply took eleven seconds to write and nine hours to arrive. Multiply it by the forty or so order-status emails a week that a store doing decent volume gets, and you have a person whose morning is spent transcribing data a computer already fetched.
That's the specific thing that sends Help Scout users looking for an alternative. The software is not the problem. The shape of it is.
What Help Scout gets right
Worth saying plainly, because a lot of comparison pages skip it: Help Scout is one of the best-built shared inboxes on the market, and it has kept a loyal customer base for over a decade by being deliberately unlike a helpdesk.
Replies go out looking like normal email. No ticket numbers, no "Case #88213 has been updated," no portal login for the customer. For a small brand where the support voice is part of the brand voice, that matters more than any feature list, and it's the single most common reason people stay.
The Shopify integration is genuinely good at what it's built for. Help Scout is part of the Shopify Plus Certified App Program, and the sidebar app surfaces order details, fulfillment status, tracking numbers, receipts, and purchase history next to the conversation, with the ability to refund, cancel, and edit orders without leaving the inbox. If you have humans answering email, this is close to the best version of that job.
Docs and Beacon round it out. You get a hosted knowledge base and an embeddable widget that suggests articles based on the page a visitor is on. Reporting is solid. Workflows handle the routing rules most teams need.
The free plan is unusually generous for the category. As of publication, Help Scout's pricing page lists a Free plan with 5 users, 1 Inbox, 1 Docs site, and 100 contacts a month, alongside a 15-day trial of Standard or Plus with no credit card. That's a real starting point, not a demo.
So a store on Help Scout usually likes it. The open question is whether an inbox is the right shape for the questions an ecommerce store gets.
Inbox-shaped, not store-shaped
Help Scout's AI and Help Scout's store data live on opposite sides of the product.
The Shopify sidebar is built for the agent. It renders in the conversation view, for a logged-in human, so a person can read it and type. The AI, meanwhile, reads something else entirely. Help Scout's own documentation is direct about it: AI Answers gets its information from the knowledge sources you add — Docs sites, publicly accessible websites, PDFs, spreadsheets — and answers only from those sources, never from the general web.
Which means AI Answers can tell a customer your return window is 30 days, because that sentence exists in an article. It cannot tell them where order #1042 is, because that fact exists in Shopify, and Shopify is not a knowledge source. Same for "is the navy medium back in stock," "does this fit a 2019 model," "what did I actually order in June." Every question whose answer depends on live store data routes to a person.
For most ecommerce stores that's the majority of the queue. Order status alone typically dominates it, and it's also the least interesting work in the building: pure lookup, zero judgment, and the customer wanted it instantly.
The knowledge-base grounding cuts the other way too, on the pre-purchase side. A shopper asking which of two jackets is warmer is asking a catalog question. If the answer isn't written up in a Docs article, AI Answers doesn't have it, and the conversation either escalates or dies. Stores lose those quietly, which is why the sales-agent side of AI support tends to be underrated next to the deflection numbers.
None of this is a flaw in Help Scout's engineering. It's a consequence of what it was built to be: a tool for helping a support team work an inbox well. Help Scout's own ecommerce page frames the value as your team not having to "open Shopify, hunt for the order, copy a tracking number, and switch back to type the reply." That's an honest description of the product, and it presumes a your-team in the sentence.
The pricing model, briefly
Help Scout bills per user. As of publication its published tiers are Standard at $25 per user per month, Plus at $45, and Pro at $75, with roughly 16% off for annual billing. AI Answers is separate: $0.75 per resolution, where a resolution means a conversation the AI closed without the customer escalating or asking for more help. Help Scout doesn't charge for a conversation where the customer hits "I still need help," and you can set a monthly spending cap. New accounts get three months of unlimited AI Answers free. (For the full tier breakdown, the inbox and Docs site add-on fees, and how the per-resolution meter behaves at volume, see our Help Scout pricing breakdown.)
Per-seat pricing suits a specific kind of team well: one where each agent handles high-value, high-judgment conversations, headcount is stable, and adding a person is a deliberate decision made twice a year. B2B software support, agencies, healthcare. For those teams $45 a seat for a well-built inbox is a fine trade.
It suits a different kind of team badly. If your ticket volume tracks order volume, and order volume swings with a product launch or Q4, then per-seat pricing prices the wrong variable. You don't need more judgment in December. You need more throughput on the same forty questions, and the only lever the pricing model gives you is hiring. Meanwhile the AI meter runs on a separate axis, so busy months move both numbers at once.
The Pro tier makes this concrete: it's $75 per user with a ten-user minimum, so the floor is $750 a month before any AI resolutions, which is the tier where SSO/SAML and HIPAA live.
Who actually outgrows it
Three profiles, from what the search traffic on this term looks like.
The two-person store drowning in WISMO. Volume is up, the inbox is full, and 60-70% of it is order status. Help Scout makes each of those replies fast, but a fast reply is still a reply. What this store needs is for those questions to never reach a human, and the AI layer it's paying for can't touch them because they're not in a Docs article.
The store where growth means hiring. Every incremental thousand orders adds hours of support work, and the response is another seat at $25 or $45. That's a support cost that scales linearly with revenue, which is exactly the curve small stores are trying to bend. Better Docs articles help at the margin, but they only deflect the questions a document can answer, and the ones driving the volume aren't those.
The store whose real problem is pre-purchase. Sizing, compatibility, stock, shipping time. These arrive on the storefront, at the moment someone is deciding whether to buy, and they age badly. An inbox tool answers them eventually. Eventually is a lost sale.
If none of those describe you, stay. A team of eight with a mixed queue, a phone line running through Aircall, SLA policies that matter, HIPAA requirements, or support that isn't ecommerce at all is on the right product already. Switching would trade a mature tool for a narrower one.
Where Ernest fits
Ernest is an AI agent for Shopify and WooCommerce stores rather than a shared inbox with AI attached. The difference that matters for the problem above: the store connection feeds the AI directly, not a sidebar. It does live order lookup and product search against the real catalog, so "where is order #1042" and "do you have this in navy, medium" get answered in the chat, at 11pm, without a person.
On connect it ingests the store's site — policies, FAQs, product pages — and imports Shopify policies and the catalog automatically. It replies in the customer's language. For returns and cancellations it can act, either fully automatically or gated behind a one-click merchant approval, your choice per action. Refunds are always approval-gated, no setting changes that.
Before it talks to a real customer, you can replay your own past tickets against it in simulation and read what it would have said. That's the honest way to answer "will this actually be right about my store," and it's a better test than any resolution-rate number a vendor publishes.
Pricing is volume-based and includes AI at every tier: free for 100 conversations that never expire (a real free plan, not a trial), then $49/month for 500 conversations, $149 for 2,000, $299 for 10,000. No per-seat fee, no per-resolution meter.
The honest cons, because the comparison is worthless without them:
- No ticketing, no SLAs, no queue management. Help Scout's workflows, views, assignment rules, round-robin routing, and SLA policies have no equivalent in Ernest. If your team coordinates work inside the inbox, that's a real loss.
- No CRM, ERP, or helpdesk integrations. No Salesforce, no HubSpot, no Jira, no returns-portal apps. Help Scout's integration directory is deep and Ernest's is two platforms.
- No phone, SMS, WhatsApp, Instagram, or Messenger. Storefront chat and email, with escalation to a human over email. Help Scout covers more channels on Standard than Ernest covers at any tier.
- No SSO/SAML, HIPAA, or IP restrictions. If procurement has a security checklist, Help Scout Pro has answers Ernest doesn't.
Some stores run both, which is a legitimate answer: Ernest on the storefront handling order status and pre-sale questions, Help Scout keeping the email queue for everything that needs a person. That drops the free-plan contact count and the seat count you need without giving up the inbox.
How to decide
Two questions settle it faster than a trial.
What fraction of your queue is answerable from a document? Go through last week's conversations and mark each one: could this have been answered from a help article, or did it need live store data? If the article column wins, Help Scout's AI is a good fit and you should tune your Docs instead of switching tools. If the store-data column wins, you're paying for AI that structurally cannot reach most of your queue.
Is your next support hire a judgment problem or a volume problem? If you need someone who can handle escalations, complaints, and edge cases, hire the person and keep the inbox. If you need the same forty answers delivered faster, that's not a headcount problem, and per-seat pricing will keep presenting it as one.
Stores weighing this against the wider field can start with our roundup of customer service software for small businesses. If the store-data column is the one that won, Shopify merchants can install Ernest from the App Store and run a month of their real conversations through simulation before changing anything.