Help Desk for a Small Online Store: What You're Actually Paying For

The support inbox breaks quietly. One month two people are answering everything from a shared Gmail account and it’s fine. The next month someone replies to a question a colleague already answered, a refund request sits unread for four days, and a customer posts about it.

Nobody decides to buy a help desk. You just reach the point where not having one costs more.

The part that catches people out

Every tool in this category will consolidate your email, chat and social messages into one place. That’s table stakes, and if that’s all you need, the cheap options do it well.

What differs — and it differs enormously — is what they charge you for.

Most help desks bill per agent. You pay a fixed amount per person who logs in, and it doesn’t matter whether they answer five tickets or five hundred. One tool in this comparison bills per ticket instead, and doesn’t charge for seats at all.

Those two models produce wildly different bills for the same shop. A store with two full-time agents and heavy volume pays one way. A store with eight part-time helpers and modest volume pays the opposite way. Same category, same job, opposite answer.

It’s the same trap I ran into comparing email tools for small stores, where some platforms charge for the size of your list and one charges for how much you send. The specifics change; the shape of the mistake doesn’t.

How I judge these tools

Four questions, same as always.

1. What does the bill look like at three times your current size? For help desks this means: three times the tickets, or three times the people? Those numbers grow at different rates, and only one of them is on your invoice.

2. What counts as billable? A seat? A ticket? A conversation? A resolution? Read this definition before you read the price, because the definition moves the number more.

3. Where does it stop? Reporting is where help desks usually run out of road. So is automation logic once your rules get conditional.

4. What does leaving cost? Ticket history exports. Macros, rules and integrations don’t. Assume you rebuild those.

How the five compare

Two things worth staring at.

Gorgias is the one tool in this table that doesn’t care how many people you hire. Everyone else multiplies by headcount.

And Help Scout has a genuine free plan — five users, one inbox — which none of the others offer. For a shop that just needs two people not to reply to the same email twice, that may be the whole answer.

What Gorgias does well

The Shopify integration is the actual product

This is the thing to understand before you look at the price. Gorgias isn’t really competing on being a better shared inbox. It’s competing on knowing what the customer bought.

Order history, refunds and cancellations sit inside the ticket. Your agent answers “where is my order” without opening another tab, and issues the refund from the same screen. On a seat-billed generalist tool you’re either switching tabs constantly or paying someone to build that bridge.

In G2 reviews, ease of use and the Shopify integration are by a distance the most-praised things about it — roughly twenty-five separate mentions each. The pattern in the praise is consistent: people describe consolidating scattered channels into one dashboard and having the commerce context already there.

Seats are free, so you can staff generously

Because billing ignores headcount, you can put your warehouse person, your part-time weekend helper and your VA into the tool without doing arithmetic first. On Zendesk Suite Team at $55 per agent, adding three occasional helpers costs $165 a month before anyone answers anything.

For shops with seasonal spikes — and most shops have seasonal spikes — this is the argument. You can hand November access to two temporary helpers and hand it back in January without the bill noticing. On per-seat pricing that same flexibility means either paying for idle seats year-round or doing a small administrative dance twice a year.

It also changes who gets to see the customer. When adding a person is free, the warehouse manager who actually knows where the parcel is can answer the question directly, instead of relaying it through whoever holds the licence.

Automation that actually reduces volume

Reviewers report meaningful automation rates, some describing well over half of tickets handled without a human. Take those numbers as directional rather than a promise, since they depend entirely on how repetitive your questions are. A shop answering “where’s my order” forty times a day automates well. A shop with complex product questions does not.

Where it falls short

The bill is hard to predict

This is the most-cited complaint in reviews, and it follows directly from the pricing model. Per-agent billing is boring and predictable: you know your headcount. Ticket volume is not under your control. A product goes viral, a shipment gets delayed, and your support bill moves with it — in the month you can least afford it.

Overage runs $0.40 a ticket on the lower plans and $0.36 at Pro and above. Three hundred unexpected tickets is an extra $120 you didn’t budget.

The AI billing has an edge worth knowing

AI resolutions are charged separately — around $1.00 each on monthly billing, less annually, with interactions beyond your bundle costing more. Here’s the part that surprises people: a conversation the AI fully resolves can count both as a billable ticket and as a paid resolution.

Conversations handed to a human revert to the normal ticket charge. So the automation that’s meant to save money is, on that specific path, billed twice. It isn’t hidden — but it isn’t what you assume either, and it should go into your model before you switch AI on.

Reporting runs out sooner than you’d like

Custom reporting is thin, and it comes up often enough in reviews to treat as a real limit rather than a preference. The shape of the complaint is consistent: people can see volume and response times, but the moment they want to ask something slightly sideways — which product generates the most refund requests, whether Tuesday tickets resolve slower — they end up exporting to a spreadsheet.

For a shop that mainly wants to know whether the queue is under control, this is fine. If support data is how you find product problems, it will frustrate you, and that frustration arrives around month two rather than on day one.

Setup takes longer than the demo suggests

Several reviewers describe a genuine learning curve, particularly around automation rules and training the AI agent. A few mention the AI closing conversations it shouldn’t have while still being tuned.

None of that is disqualifying — it’s what configuring any rules engine looks like — but it does mean this is not a tool you switch to on a Friday and have working by Monday. Budget a proper week, and don’t schedule the migration anywhere near your busiest month.

The alternatives, and when to pick them instead

Help Scout — pick this if you have few tickets and a small team. The free plan covers five users and one inbox, which is more than a lot of shops need. Standard is $25 per user, Plus is $45, Pro is $75, and the 15-day trial doesn’t ask for a card. AI Answers is a separate $0.75 per resolution. It won’t know what your customer ordered, and for many shops that genuinely doesn’t matter.

Freshdesk — pick this if you want per-agent billing at the lowest price. Growth is $19 per agent annually, Pro $55, Enterprise $89. It’s the cheapest per-seat option here by a clear margin. The 14-day trial drops you into the Enterprise tier, which is generous for evaluating and slightly misleading about what you’ll actually live with.

Zendesk Suite — pick this if you’re heading somewhere large. Team is $55 per agent annually, Professional $115, and monthly billing costs meaningfully more than annual. There’s no free plan. You’re buying depth and a very large integration ecosystem, and paying a premium per head for it. For a shop with three agents this is hard to justify; for one with thirty it starts to make sense.

Re:amaze — pick this if you have many hands and modest volume. It’s the interesting middle. Per-user plans run $29 / $49 / $69, but the Starter plan is $59 flat with unlimited team members, capped at 500 responded conversations a month. If your problem is “eight people need access and none of them are full-time,” that single plan may be cheaper than anything else in this table.

What it actually costs

Two shops, same category, opposite answers.

Shop A: two full-time agents, 1,500 tickets a month.

Gorgias is roughly nine times the cheapest option. Ticket volume is doing all the damage, and the two free seats you’re not paying for don’t come close to covering it.

Shop B: eight part-time helpers, 400 tickets a month.

Now the ordering inverts almost completely. Zendesk is four times Gorgias, and the two volume-priced options sit at the top.

So the question is simply: which is growing faster, your ticket count or your headcount? If tickets are climbing and your team is fixed, per-agent billing protects you. If you keep adding occasional helpers and volume is flat, per-ticket or flat-rate billing protects you.

And there’s a third case worth naming honestly: if what you actually need is the commerce integration — refunds and order lookups inside the ticket — then this comparison of monthly figures is beside the point. You’re buying something the seat-billed tools don’t sell, and the question becomes whether it saves more agent time than it costs.

One more line item that moves the totals: annual billing takes roughly 16% off the Gorgias figures above, and the per-agent tools discount annually too — Zendesk’s monthly price runs meaningfully above its annual rate, and Freshdesk advertises around 20% off for paying yearly. Every number in the tables above is the monthly-billing case, which is the honest one to compare when you don’t yet know whether you’ll still be using the tool in a year.

Prices move, and each company’s pricing page is the authority. Check before you commit.

Who it’s for, and who should skip it

The Shopify store where agents live in the order screen. If your tickets are mostly “where is it”, “can I change the address”, “please refund this”, the integration pays for itself in saved clicks. This is the case Gorgias is built for, and the one where the price makes sense.

The two-person shop with a busy inbox. Skip it. At 1,500 tickets and two agents you’d pay $360 where Freshdesk charges $38. Unless the commerce integration is saving you hours a week, that gap is very hard to argue with.

The shop that needs deep custom reporting. Skip it. Thin custom reporting is a recurring complaint, and if analysing conversation patterns is central to how you run support, you’ll be exporting to a spreadsheet within a month.

What to test in the first week

Trials get wasted on feature tours. Three things actually decide this one.

Put a real week of tickets through it. Not sample data — your actual queue, including the awkward ones. The point is to see your ticket count land somewhere in the plan tiers, because that number is your bill.

Time one refund end to end. Open a genuine “I want to return this” ticket and complete it. If the order data and refund control being inside the ticket saves you thirty seconds, multiply that by your monthly volume and you have the real argument for paying more than a per-seat tool charges. If it saves nothing because you were never in the order screen anyway, you have your answer too.

Ask one awkward reporting question. Pick something you’d genuinely want to know — which product drives the most complaints, say — and try to answer it without exporting. Whether you can is the clearest signal of whether this tool fits how you work.

Common questions

Is there a free plan? No. There’s a trial, and the entry Starter plan is $10 a month for 50 tickets — cheap enough to test with real traffic. If you want a permanently free option in this category, Help Scout’s free tier covers five users and one inbox.

How does per-ticket billing actually work? Each plan includes a monthly ticket allowance — 50, 300, 2,000, 5,000 depending on tier — and tickets beyond it cost $0.40 each, dropping to $0.36 at Pro and above. Seats are unlimited and not billed.

Does the AI cost extra? Yes, separately from your plan. Around $1.00 per resolved conversation on monthly billing, less on annual, with interactions beyond your bundle costing more. A fully automated resolution can be billed as both a ticket and a resolution, so model that before switching it on.

Is it cheaper than Zendesk? It depends entirely on your shape. With eight helpers and 400 tickets, Gorgias is roughly a quarter of Zendesk. With two agents and 1,500 tickets, it’s more than triple. Count both numbers before deciding.

How hard is it to leave? Ticket history exports. Macros, automation rules and integrations generally don’t travel, so budget time to rebuild them. This is true across the category, not specific to one tool.

The bottom line

Gorgias makes sense for a store whose support is really commerce work — order lookups, address changes, refunds — and whose team is wider than it is busy. Free seats and the Shopify integration are the two things it has that the per-agent tools don’t.

If you have a small team answering a lot of tickets, it will cost you several times what a per-seat tool charges, and no amount of feature comparison changes that arithmetic. That’s the honest headline.

The useful test is cheap: count your monthly tickets and count the people who need access, then put both numbers into the tables above. If the ticket number is the big one, start your trial somewhere else. If the headcount number is the big one, the Starter plan is $10 for a month — run a week of real tickets through it and see whether the order data inside the ticket changes how fast your team works.