Written by an Adobe Commerce engineer, not a review site

Your contact form is the most expensive page on your store

I install live chat on Adobe Commerce, BigCommerce and WooCommerce stores for a living. This is the honest comparison — including the two situations where the tool I'm paid to recommend is the wrong answer.

Disclosure: I'm a LiveChat partner and I earn a commission if you sign up through my link. I install it on client stores either way — that's how I formed the opinion. Where a competitor is the better fit, this page says so.

The number that started this

Gorgias published support benchmarks across more than a thousand ecommerce brands, split by the channel a conversation started in. The spread is not subtle.

13%abandon the conversation when it starts in live chat8h median resolution
30%abandon when it starts by email32h median resolution
42%abandon when it starts on a contact form36h median resolution

Same customer. Same question. Three times the loss — and the only variable is the container you put the conversation in.

To be precise about what that 42% is: it's conversation abandonment, not cart abandonment. Forty-two percent of the people who fill in your contact form give up before you answer them. That's not the same claim as "42% of your carts", and I'm not going to make the bigger one.

What makes it expensive rather than merely annoying is which questions arrive there. On the stores I audit, 70–80% of contact-form submissions are pre-purchase: does this fit a 2019 Tacoma, is it really in stock, can I return it if the sizing is wrong, will it ship to Alaska. Those aren't support tickets. They're sales conversations arriving in a queue with a 36-hour SLA.

Baymard Institute's data on why shoppers abandon carts — excluding the ones who were only browsing — puts 19% on not trusting the site with their card details, 13% on the returns policy, and 12% on not being able to see the total cost up front. Every one of those is a forty-second answer from a human who happens to be present. Grouping them as "problems a conversation solves" is my interpretation, not Baymard's finding — but I don't think it's a stretch.

Who this page is for

Most comparison pages try to sell everybody. This one is trying to lose about two-thirds of you, because a recommendation that fits everyone fits nobody.

Keep reading if

  • You're on Adobe Commerce, Magento, BigCommerce or WooCommerce — where the ecommerce-native support tools barely show up
  • Your AOV is high enough that a chat is a sales conversation: parts, furniture, jewellery, industrial, B2B wholesale
  • You run 2–10 full-time agents, not a rotating seasonal pool
  • You've had an overage invoice from Zendesk or Gorgias and didn't enjoy it
  • You need SSO, SLAs or data-region control

Close this tab if

  • You're on Shopify with high volume and low AOV — Gorgias is your answer and it isn't close
  • You're under about $250k — Shopify Inbox and Tawk.to are free and genuinely fine at that stage
  • You run many part-time or seasonal agents — per-seat billing will punish you
  • Your stated goal is "deflect 60% of tickets with AI" — you need an agent that can process a refund, and that's not this

The pricing question nobody asks properly

The category has quietly repriced itself. Gorgias charges around $0.90 per AI resolution. Intercom Fin is $0.99. Zendesk is roughly $1.50 committed and $2.00 pay-as-you-go, on a meter that auto-charges overages. Text's own newer plans meter AI at $0.99 too.

Which means your support cost is now a function of your traffic. Black Friday gives you four times the volume and four times the invoice — in the week your margin is thinnest, because you discounted everything, paid for the ads and ate the shipping promo. I have yet to meet a merchant who modelled this before signing.

Flat per-seat is the counter-position: ten thousand chats and a hundred thousand chats cost the same. It loses badly if you run a large team on low volume. It wins decisively if you run a small team on high volume — which describes most ecommerce.

Run your own numbers

No email required. This is the same calculation I run for clients, and it sometimes tells them to buy the other thing.

Per-resolution model$0
Flat per-seat @ $79/agent$0

Defaults: 45% is the median AI resolution rate Gorgias reports across its merchant base; $0.99 is Intercom Fin's and Text's published rate; $79 is LiveChat's Business tier on annual billing. Change all four to match your reality — and check whether your current contract's meter is capped. Several aren't.

The honest comparison

ToolPick it whenWhere it hurts
Gorgias Shopify, high volume, low AOV, and you want AI that can genuinely take actions on orders — refunds, cancellations, address changes. Effectively Shopify-or-bust. Your bill scales with traffic and peaks in Q4.
Intercom (Fin) You're a SaaS or subscription business more than a store, and you need product tours and onboarding alongside support. $0.99 a resolution, and it's built for software companies, not catalogues.
Zendesk You're already on it enterprise-wide and switching costs are real. The most expensive per-resolution rate in the category, on an uncapped meter.
Shopify Inbox / Tawk.to Under ~$250k in revenue. Free is the correct answer and anyone telling you otherwise is selling. Limited routing, reporting and integration depth once you have a real team.
LiveChat
MY RECOMMENDATION
Adobe Commerce, BigCommerce or WooCommerce; 2–10 full-time agents; AOV high enough that chat is a sales channel; and you want a support bill you can forecast twelve months out. Flat per-seat on legacy plans. 100% uptime across all services June–August 2026. Three real weaknesses. Per-agent pricing gets expensive past ~8 seats. Its AI suggests replies but can't take actions on an order. And it's a minor player in the Shopify ecosystem — 43 reviews on that app store against Shopify Inbox's 5,500.

If the middle column of that last row isn't describing you, one of the other four is your answer. I'd rather you got that right than clicked my link.

What I'll do if you sign up through me

Free installation and configuration on your store

A chat widget takes thirty seconds to paste in and about ninety minutes to do properly. The gap between those two numbers is where most implementations quietly fail — and it's the part I actually do for a living.

  • Loaded correctly — deferred, async, after first paint, and never through GTM if you care about LCP. On a Hyvä theme this is the difference between a Lighthouse hit you can ignore and one you can't.
  • Product context wired in — SKU, price and stock status passed into the visitor payload, so your agent isn't asking a question the page already answered.
  • Routing by category, not round-robin — the person who knows outboard motors shouldn't be fielding questions about anodes.
  • Checkout handled deliberately — present, but minimised and without a proactive greeting. A popup over a payment form is a bounce, not a conversation.
  • Assisted revenue tracked from day one — the only number that ends the "is this worth it" argument. Without it you'll cancel in month three for no reason.
Start a free LiveChat trial →

No credit card required · Commissioned link · Then email me for the install

Not sure it's the right fit? Send me your store URL and I'll run a six-point audit of your pre-purchase support path and send back a short Loom. No charge, nothing attached — I look at a lot of stores and the patterns are useful to me. If the audit says you need Gorgias, it'll say Gorgias.

Sources

Every figure on this page traces to a named organisation with a published methodology. The live chat industry runs on statistics nobody can source; these are the ones that survive being checked.