Reduce Support Ticket Volume Without Annoying Customers

Most support teams deal with a growing queue by answering faster. That works for a while. The cheaper fix is to stop the same messages from arriving at all, because in almost every queue a large share of the work is a handful of questions asked again and again.

To reduce support ticket volume without making customers feel shut out, find out why people write, remove the reasons you can, and make the rest quick to answer. This guide goes through it in that order, with a spreadsheet tally and a worked example for measuring the result.

Cutting volume is not the goal by itself

Hiding the contact button cuts volume too. The problems then turn up as one-star reviews and chargebacks. The volume worth removing is the contacts that happen because something on your side was unclear, broken or late. Someone asking for advice before buying is not volume to get rid of.

One test for every change below: the customer's problem goes away, and reaching you stays exactly as easy as before.

Step 1: find out why people contact you

If conversations carry labels and tickets have categories, export the last four weeks and count. Labels are often patchy, though, so for the first round a hand count of a sample is more reliable.

A tally you can do in a spreadsheet

1. Pick 200 recent conversations and tickets from all channels, from an ordinary month (no launch week, no sale).
2. Make four columns: ID, contact reason, root cause, preventable (yes, partly, no).
3. Read the first message and the final answer. Write the reason the way the customer would say it: "where is my order", not "logistics inquiry".
4. After 50 rows, merge similar reasons into 8 to 12 groups and fix the earlier rows.
5. Count each group with a pivot table or =COUNTIF(B:B,"where is my order"), then sort.
6. Divide each count by 200 to get its share.

An invented tally from an online shop:

The top three preventable reasons are 59% of the sample (29 + 17 + 13), and order status alone is 29%, so that's where this shop starts. The pre-sale row is 15% of the queue and the kind of conversation you want more of.

Turn the groups into labels so the next count comes from your reports. The knowledge base guide uses a similar count to choose which help articles to write first.

Step 2: fix the cause with the team that owns it

Support can't fix most top reasons alone. Bring the team that owns each one the numbers and a few real conversations with personal data removed. A one-page note gets read; a dashboard link usually doesn't.

• Reason: customers can't find their tracking number
• Volume: 58 of 200 sampled (29%), about 87 a week at our current 300 a week
• What customers say: "I paid on Monday, where is my package?" (three more examples attached)
• Likely cause: the shipping email only goes out when the courier scans the parcel, often two days after payment
• Proposed fix: an "order packed" email on day one with the expected dispatch date
• How we'll know: order-status contacts per 100 orders, weekly, for four weeks

Ask for small fixes first, like a reset link that doesn't expire after ten minutes. Their numbers make the case for bigger changes.

Order and status emails

For shops, the order emails are usually the biggest lever. Each one should answer the next question: the confirmation says when the order will be packed, the dispatch email has a working tracking link, a delay email goes out before the customer notices, and the refund email says how many days the bank takes. Say in every email where to reply, or people open a chat and the agent starts with no context.

Step 3: self-service for what can't be designed away

Return windows, sizing, changing a delivery address: some questions stay. Write each one up as a short help page or FAQ entry on your site. Hodhod does not host a help center, so you publish the pages on your own site or in a docs tool and link them from conversations, canned replies and bot options.

Self-service only cuts volume if people see it before they write, so put the link on the tracking page and in the order email.

Step 4: tell customers before they ask

When something breaks, the queue fills with one question. Two Hodhod widget features are made for this:

• Widget announcements: up to two notices at the top of the widget per website inbox, a yellow "announcement" or a red "warning", with an optional start and end time. Good for things you know in advance, like a holiday closure or slower shipping during a sale. More in announcements and warnings in the widget.
• Issue notice: when the team tracks an ongoing problem as an issue, the widget can show a notice such as "The payment gateway is being looked into". It's off by default and turned on per inbox. After the fix you can send a one-time message to the open conversations linked to the issue.

Step 5: make the repeat answers fast

Write a canned response for each of your top ten reasons and review them monthly, since a wrong canned answer gets sent a hundred times. Then add a decision-flow bot to the widget with options like "Track my order", "Returns" and "Something else". Each option gives the answer with a link, hands over to an agent, or opens a ticket. Always keep a way to reach a person; a bot with no exit only moves the problem to reviews. Once a month, open the flow's "Handoffs" report to see at which step people most often ask for an agent, and rewrite that step.

Step 6: choose when chat is offered

Each website inbox has a contact mode: live chat only, ticket only, both, or ticket outside business hours. The last one doesn't stop people writing, but it removes the "hello?" and "is anyone there?" follow-ups and the second chat from the same person at midnight. The customer leaves one ticket with details and gets the answer by email. The live chat page describes each mode.

Step 7: measure the effect over several weeks

Track contacts per 100 orders (or per 100 active users for software), not raw volume, because volume moves with sales. Compare four weeks before with four weeks after, leaving out the two weeks right after the changes while things settle.

1,200 / 8,000 × 100 = 15, and 1,080 / 9,000 × 100 = 12. Raw volume fell 10% (120 of 1,200), but the rate fell 20% (3 of 15) because orders grew at the same time. At the old rate, 9,000 orders would have brought 9,000 × 0.15 = 1,350 contacts. The team got 1,080, which is 270 fewer in four weeks.

Then check that customers didn't pay for the drop:

• Satisfaction scores after conversations stayed level or rose.
• Repeat contacts about the same problem within seven days did not rise.
• Pre-sale conversations held steady.
• Reviews and refund requests did not increase.

Repeat the tally each quarter. Reasons shift with the product and the season.

To try labels, widget announcements and a support bot on your own traffic, request a demo. Every plan starts with a 14-day free trial.

Frequently asked questions