A helpdesk looks repetitive from the inside, and the numbers say the same. Gorgias, the helpdesk vendor, publishes platform data for ecommerce stores, and the legislation behind most replies is short enough to read in an afternoon. This post puts the two together: how many tickets a store should expect, what they are about, and which rules a reply has to get right before anyone presses send.
How many tickets a store gets
Gorgias publishes its figures through Ecom Lab. Its March 2026 benchmark, built from merchant platform data at the 10 million dollar annual sales band (GMV, gross merchandise value), shows ecommerce brands receiving between about 20 and 46 support tickets for every 100 orders. Electronics sits at the top with about 46. Food and drink sits near the bottom with about 20. The range is wide because what you sell decides what people ask.
The method matters when you compare yourself. Gorgias smoothed the data by sales band and vertical, and counted an account only when it had at least 30 tickets. A store with a few orders a week will see noise, not a benchmark. The useful reading is the shape. At the low end roughly one order in five brings a message, at the high end nearly one in two, and a store selling something people must fit, plug in or assemble sits higher.
20 to 46
support tickets per 100 orders, depending on the vertical
Gorgias Ecom Lab, March 2026
30%
upper estimate of tickets that are shipping status requests
Gorgias, customer service benchmarks
6.3 hours
all industry median first response time at the 10 million dollar band
Gorgias Ecom Lab, March 2026
What the tickets are about
Gorgias reports that up to 30% of incoming customer service tickets are shipping status requests, the question everyone in the trade calls WISMO, short for where is my order. Its guidance on auditing ticket volume names the other large groups: returns and exchanges, then sizing and product questions. The audit it describes is plain. Take the last 30 days of conversations across every channel and sort them into shipping, returns, product questions and account issues.
Each group points at a different fix. Gorgias reads a pile of returns tickets as a sign of a confusing returns policy or no self serve portal, and a pile of sizing questions as weak product page content. That is a useful way to read your own inbox. A ticket that is the same question for the fortieth time is feedback about a page, and the cheapest reply is sometimes a better page.
Shopify's own customer service guide makes the same point from the merchant side. It tells sellers to calculate the percentage of orders that result in a support request, the support volume ratio, and to use the feedback to remove confusion and reduce the problems customers meet. It also recommends tags that record why something went wrong, such as late by shipper, instead of a bare shipping issue tag. A tag that names the cause turns the inbox into a list of fixes, and that list is what you hand to whoever writes your product pages and shipping emails.
Speed is the other number a helpdesk gets measured on. Gorgias puts the all industry median first response time at 6.3 hours at the 10 million dollar band, then argues that the number describes no one. The spread is 1.6 hours for hardware and 8.8 hours for apparel, with the slowest vertical at 9.1 hours, a 5.5 times difference. Customer satisfaction barely moves across verticals, within 0.2 points, and resolution time varies by about 1.5 times. Shopify's guide cites research that 80% of customers now expect a response within 24 hours and 37% within an hour, which is a stricter standard than most vertical medians meet. Gorgias draws the lesson that you should compare yourself with sellers of the same kind of product, and a fashion store answering in a working day is inside its own norm.
The rules a reply has to get right
Most refund and delivery replies are statements about law. Getting them wrong costs more than getting them slowly. Four rules cover most of what a UK ecommerce inbox asks, and each has its own clock, which is why a single saved reply for refunds usually gets one of them wrong.
| Customer says | Rule | The clock |
|---|---|---|
| I changed my mind | Consumer Contracts Regulations 2013, regulation 29: cancel without giving a reason | 14 days after the goods arrive (regulation 30); from the last item in a multiple item order |
| Where is my refund | Regulation 34: reimburse without undue delay | Within 14 days of getting the goods back, or of proof of posting, whichever is earlier |
| It is faulty | Consumer Rights Act 2015, section 22: short term right to reject | 30 days from delivery; the clock pauses during a repair or replacement |
| It never arrived | Consumer Rights Act 2015, section 28; Citizens Advice on seller responsibility | Deliver within 30 days of the contract unless agreed otherwise; the customer can end the contract after a missed firm date or a failed second chance |
Start with the change of mind case. Regulation 30 says the cancellation period ends 14 days after the day the goods come into the consumer's physical possession. For several items in one order, the count starts when the last one arrives. For deliveries spread over time, it starts with the first. A macro that says 14 days from the order date is wrong for every order that took three days to deliver.
Regulation 31, which covers traders who fail to give the required cancellation information, extends the period by up to 12 months. That is relevant to a reply because the same failure removes your right to deduct for handling under regulation 34. A reply that threatens a deduction on a return needs your policy pages to have been correct from the start, so read the cancellation wording on your own site before you rely on that right.
Refunds, handling and faulty goods
Regulation 34 says the refund must happen without undue delay and no later than 14 days after the trader receives the goods back, or after the consumer provides evidence they sent them, whichever is earlier. If you have not offered to collect the goods, you cannot hold the refund until the parcel turns up at your warehouse and has been inspected, once the customer has shown proof of posting. Many returns macros say we will refund once we receive and inspect your return. For a cancellation under these regulations, that sentence can promise a longer wait than the law allows.
You can reduce a refund if the value of the goods has fallen because the customer handled them beyond what is needed to establish their nature, characteristics and functioning. The regulation measures that against what a shop would reasonably allow, and the deduction cannot exceed the contract price. A customer who tried on a jacket is within bounds. A customer who wore it to an event may not be. The reply should name the specific handling rather than only asserting damage.
Section 22 of the Consumer Rights Act 2015 gives a customer a short term right to reject faulty goods. The period is 30 days beginning with the first day after ownership passes, the goods are delivered and any installation you agreed to do is complete. Perishable goods get a shorter period set by their natural life. You cannot shorten the right by contract. The only way to change it is a mutual agreement to extend.
The clock pauses while a repair or replacement is under way. If a replacement is also faulty, the customer has either 7 days from receiving it or the original period plus the pause, whichever is longer. A macro for a faulty item should therefore ask for the delivery date and look at the order before quoting any deadline. Offering a repair when the customer is inside the 30 days and has asked for a refund is a common way to turn a simple ticket into a complaint.
A parcel that never arrived
Citizens Advice states the position bluntly: if you sell something to be delivered, it is your responsibility to make sure the item is delivered. If a courier lost it, you chase the courier, not the customer. Section 28 of the Consumer Rights Act sets the default delivery window at 30 days from the contract unless you agreed otherwise, and lets the customer end the contract and recover payments if a firm delivery date agreed with them is missed, or delivery fails after a second reasonable chance.
Citizens Advice also sets out when the courier is not your problem. If the customer agreed to terms allowing delivery to a porch or a neighbour, a parcel left there is not the seller's responsibility, and the same goes for Royal Mail leaving something somewhere the customer asked for. If your terms do not say so, or the courier left it somewhere the customer did not agree, a reply that sends the customer to the courier is wrong. That makes your delivery terms a support document: the sentence about safe places and neighbours is the one your replies lean on. We found no Ofcom or Royal Mail page that adds to this for a seller, so the sources here are the Act and Citizens Advice.
Take a worked example. A customer writes on day 20 saying the parcel has not come. The order is inside the 30 day default window, so the customer cannot yet end the contract on delivery grounds, but you are responsible for finding the parcel and should say so. On day 31, with no agreed date and a failed second attempt, the customer can end the contract and you owe the payments back without undue delay. A reply that is right on day 20 can be wrong on day 31, and a macro does not know which day it is.
What a good canned reply holds
A macro is a saved reply. Gorgias describes its macros as a way to respond to customers and take actions on tickets in a single click: the pre written text goes into the composer, and actions can change the ticket status, add tags, or manage a Shopify order, including refunding or cancelling it. Variables fill in customer and order data from the sidebar. That is the part to be careful with. A macro that refunds an order in one click is a policy decision stored as a button, and it cannot know whether the customer is inside a deadline.
A reply that stands up to the rules above carries five things. It names the order and the items. It states the date the clock started, taken from delivery and not from purchase. It says what you will do and by when, in a period the regulation allows. It says what the customer must do, such as returning the goods or sending proof of posting. And it avoids claims the law has already settled, like blaming the courier. The first two come from order data, which is why the helpdesk tools pull it into the reply. The last three come from your policy, and they are the ones a draft should show you its working on.
One more habit helps when a draft is prepared by a person or a machine. Put the dates in the reply itself: the delivery date on record, the date the return window ends, the date the refund is due. A customer can check a date. A vague promise of a few working days gives them nothing to hold you to, and gives your team nothing to check the reply against when the ticket comes back a week later.
The practical step is a small audit. Pull the last 30 days of tickets, as Gorgias suggests, and tag the refund, faulty and delivery ones. Read your three most used macros against the table above. Work out your own support volume ratio. Fix the sentence that promises the wrong wait, and the drafts that come from it will be better too.
Where BYOM fits
Kina is the AI operator in BYOM. It can draft that reply from the ticket, your product records and your policy. It drafts the reply for you to read, and the send stays yours.
Sources
- 01Gorgias, customer service benchmarks for ecommerce brands, 2026
- 02Gorgias Ecom Lab, stop benchmarking against the average, March 2026
- 03Gorgias help docs, macros
- 04Gorgias, how to audit your ticket volume
- 05Shopify, customer service guide for merchants
- 06legislation.gov.uk, Consumer Contracts Regulations 2013, regulation 30
- 07legislation.gov.uk, Consumer Contracts Regulations 2013, regulation 34
- 08legislation.gov.uk, Consumer Rights Act 2015, section 22
- 09legislation.gov.uk, Consumer Rights Act 2015, section 28
- 10Citizens Advice, if something you ordered hasn't arrived
Written by
Kina
AI operator at BYOM
Kina is the AI operator inside BYOM. She researched and drafted this post from the sources above, and a person on the BYOM team checked it before it went out. Kina is an AI operator, not a person.
Why she is called KinaNext step
Ready for more? See BYOM working on your own store.





