Tawk.to Alternative for Shopify Stores: Free Chat Still Needs Someone Answering It
Tawk.to's live chat is free for life, but free chat still needs a human in it. A Tawk.to alternative for stores that can't staff the chat queue overnight.
It's 11pm on a Friday. The Tawk.to widget on your store is installed, working, and costing you nothing, which is exactly what it promised. What it isn't doing is answering the shopper who wants to know whether the medium runs small, or the customer asking where order 3312 is. By Monday morning there are fourteen unread chats. Most of them are order status. Two of those people also emailed. One of them bought the same thing somewhere else on Saturday.
That weekend queue is what sends store owners looking for a Tawk.to alternative, and the honest starting point is that the software isn't what went wrong.
The free part is real, and it's unusual
Tawk.to has been free since 2013, and the founder's own explanation of why sits on their pricing page rather than in a blog post somewhere. The monetisation is at the edges. As of publication, removing the small "Powered by tawk.to" branding costs $29/month, and their page is explicit that if you leave the branding in place, the product is free for life. There's no fourteen-day clock on it and no seat count that ends it.
The paid extras, again as of publication and all from tawk.to's own pricing page:
- Branding removal: $29/month
- AI Assist, their chatbot that answers from your knowledge base and shortcut FAQs: from as little as $29/month, with what they describe as a forever-free plan for sites with limited chat volume
- Hired chat agents, trained on your business and answering on your behalf 24x7: $1/hour
- Virtual assistants for design, admin, HR, bookkeeping and data entry: from $7/hour
The surrounding product list isn't a stripped demo either. Their page describes the hosted knowledge base and the ticketing tool as free outright, alongside chat pages for businesses without a website and a contacts and CRM view. In-chat payments are listed in alpha, and Messaging (Facebook, Google RCS, Telegram, SMS and social) and Phone are both listed as beta.
The scale claims on that page are worth quoting because they're the reason so many stores land on Tawk.to in the first place. Citing Datanyze and BuiltWith, they put themselves at 20.07% of the live chat market, ahead of Zendesk at 12.25% and Tidio at 5.44%, out of 206 tracked competitors. They say 1.7 billion people interact with a tawk.to widget every month.
If what you want is a chat box on your storefront and you have somebody who can sit in it, install Tawk.to and stop reading. None of what follows is an argument against that. It's a good deal, and pretending otherwise to sell you something would be an insult to anyone who has actually compared the invoices.
The cost that was never the subscription
Tawk.to states the real problem more bluntly than most of their competitors would. From their own page: the cost of the software isn't the biggest barrier to offering live chat, "it is the cost of staffing to answer chats 24×7 that is so prohibitive."
They then work the number. At $15/hour for a decent agent, covering chat 24 hours a day, seven days a week comes to $131,400/year for one agent's worth of coverage. Against that, their hired-agent service at $1/hour totals what their page puts at $8,736/year. It's a persuasive comparison and the arithmetic isn't the trick, because the underlying point holds: for most small businesses, the chat software was never the expensive part.
So the question for a store is what happens during the hours nobody is in the inbox, and which of the available answers fits the kind of questions that store gets.
Ecommerce chat is mostly two questions
Run a week of your own chat transcripts and the shape is usually the same. Where is my order, and can I return this. Sizing, stock and shipping cost take most of the rest. That distribution matters when you're choosing what fills the seat, because those questions split cleanly into two kinds.
Sizing, shipping policy, materials, care instructions and return windows are all questions whose answers live in documents. A knowledge base can hold them, which is exactly the model Tawk.to's AI Assist is built on, and it's the model their hired agents work in too. The founder's argument for third-party agents is that an agent doesn't need to remember every answer, only where to find it, and for document-shaped questions that's true.
Order status is a different kind of question. The answer isn't in a document, it's in a record that changes hourly, tied to one specific customer who has to be identified before you hand it over. Answering "where is order 3312" means looking it up in Shopify or WooCommerce, checking it belongs to the person asking, and reading back the current fulfillment state. An outsourced agent can do that if you give them admin access and they're willing to work in your store's back office. A knowledge base bot cannot, because the answer was never in the knowledge base.
For most stores, that one question type is the largest slice of the queue. It's also the slice that generates the follow-up email when it goes unanswered, which is how one question quietly becomes three touches. If you want the wider version of that argument, we wrote about how to reduce support tickets by fixing the questions that repeat.
What Tawk.to doesn't publish
Being precise about the gaps matters more than being dramatic about them.
Their pricing page prices AI Assist at "from as little as $29/month" and mentions a forever-free plan for sites with limited chat volume. It doesn't publish where that free volume limit sits, or what the tiers above $29/month cost as volume grows. If AI is the part you're actually buying, that's the number to get in writing before you commit.
Same with the hired agents. The $1/hour figure is stated plainly. Minimum commitments, how hours are purchased, coverage guarantees and what happens when an agent hits a question outside their training aren't on that page.
And their published product list doesn't describe a Shopify or WooCommerce order lookup. That is not the same sentence as "Tawk.to can't do it" and I'm not going to write the stronger one. If live order data inside the chat is what you're shopping for, check their integration docs directly rather than trusting anyone's summary, including this one.
Where Ernest fits
Ernest is an AI agent built specifically for Shopify and WooCommerce stores. It plays three roles from one connection: support (order status, returns, cancellations, shipping and policy questions), sales (sizing, fit, stock, shipping cost, which product to buy), and product (specs, materials, care and how-to). On connect it ingests your site content, FAQs and policies, and imports your Shopify catalog and policies automatically.
The difference that matters against a knowledge-base bot is the order lookup. Ernest queries live Shopify order data, so "where is 3312" gets a real answer at 2am rather than a promise that someone will check on Monday. Returns and cancellations can run fully automatic or require one-click merchant approval, whichever you set per action. Refunds are always approval-gated, with no setting that changes it. Anything it can't handle escalates to a human by email with the full conversation attached, which is the handoff pattern we wrote about here.
Pricing is volume-based and it's one number rather than a base plan plus add-ons: free at $0 for 100 conversations that never expire (a real free plan, not a trial), then $49/month for 500 conversations, $149/month for 2,000, and $299/month for 10,000. AI is included in every tier, including the free one. No per-seat fees and no per-resolution fees.
The limits, stated plainly, because a comparison that skips them is worthless:
- Fewer channels. Ernest covers the storefront widget and email, with escalation to a human by email. No phone, SMS, WhatsApp, Instagram or Messenger. Tawk.to's own product list shows Messaging and Phone in beta; if consolidating social channels is the job, Ernest isn't the tool.
- No hired humans. There is no equivalent to the $1/hour agent service. Ernest answers what it can and escalates the rest to you. If what you want is a person to take the queue off your hands entirely, Tawk.to sells that and we don't.
- No free unlimited human chat. Tawk.to's free live chat has no conversation ceiling at all. Ernest's free plan stops at 100 conversations, lifetime. Past that it's a paid plan, and for a store with a person who can staff chat all day, free-forever is straightforwardly the better economics.
- Refunds always need you. If you want an agent that can issue refunds unattended, Ernest will never do that.
How to decide
Two questions settle this without trialling anything.
Do you have somebody who can sit in the chat? If yes, and their hours reasonably match when your customers shop, Tawk.to's free tier is hard to argue with and the $29/month branding removal is the only line item you'll likely want. Adding AI on top of that is a separate decision you can make later.
What share of your chats need a lookup rather than a document? Pull last month's transcripts and count. If most of it is policy, sizing and shipping questions, a knowledge-base bot covers you and Tawk.to's AI Assist starts at $29/month against Ernest's $49/month for 500 conversations. If most of it is "where is my order" and return requests against specific orders, a bot that can't reach your store's order records will keep escalating the majority of the queue back to you, and the free chat window ends up costing you the weekend regardless.
Stores comparing more than one option on this axis may find the Tidio alternative comparison useful too, since Tidio's free plan splits the chat bill and the AI bill in a way that catches people out for a different reason.
If order lookups are the bulk of your queue, Shopify stores can install Ernest from the App Store and replay their own past tickets against it in simulation before a single live customer talks to it. That answers the accuracy question with your own transcripts rather than anyone's marketing copy.