Most help desk dashboards show 20+ metrics, but a store handling 150 tickets a month needs about six. These customer support metrics for small business are first response time, time to resolution, resolution rate, reopen rate, CSAT, and tickets per 100 orders. Everything else is decoration at your volume.
Key takeaways
- Track six numbers: first response time, time to resolution, resolution rate, reopen rate, CSAT, tickets per 100 orders.
- At 150 tickets a month, one angry customer moves any percentage by several points.
- Tickets per 100 orders usually lands between 5 and 15 for healthy stores.
- A spreadsheet handles four of the six metrics.
What should a small store track?
A small Shopify store should track six Shopify customer support KPIs: first response time, time to resolution, resolution rate, reopen rate, CSAT, and tickets per 100 orders. Together they cover speed, outcome, quality, and demand. Reviewing them takes about 15 minutes a month.
Here is what each one tells you:
- First response time: how long a customer waits before a human or AI acknowledges them. It drives the first impression.
- Time to resolution: how long until the issue is actually closed. This is the number customers feel.
- Resolution rate: the share of tickets closed within a set window, such as 7 days.
- Reopen rate: the share of closed tickets where the customer writes back about the same issue. It exposes fake resolutions.
- CSAT: the percentage of survey respondents who rate the interaction 4 or 5 out of 5.
- Tickets per 100 orders: support demand normalized for sales volume.
The last one is the most underrated. Raw ticket counts rise when sales rise, which tells you nothing. A ratio tells you whether your product pages, shipping updates, and policies are working.
First response time vs resolution time: which matters more?
First response time measures the wait until a first reply; time to resolution measures the wait until the problem is closed. For most small stores, resolution time drives satisfaction more, but a first reply within 4 hours stops customers from sending a second and third message.
The two numbers diagnose different problems. Slow first response means a staffing or notification problem: nobody saw the email. Slow resolution with fast first response means a process problem: you reply quickly, then wait on a carrier, a supplier, or an internal decision.
Use medians, not averages. One refund dispute that drags on for 19 days will wreck a monthly average built from 150 tickets. The median shows what a typical customer experienced.
Watch out for the auto-acknowledgment trap. An automatic 'we got your message' email technically counts as a first response in many help desks. Measure the first substantive reply, or your number will look excellent and mean nothing.
What does each metric look like in practice?
The table below gives the formula, a realistic target for a store handling 100-500 tickets a month, and where to find the data. Targets are working ranges based on common small-store practice, not industry law. Adjust them to your product and margins.
| Metric | Formula | Target range (100-500 tickets/mo) | Where to find the data |
|---|---|---|---|
| First response time | Median time from ticket created to first substantive reply | Under 4 business hours (email); under 5 minutes (chat) | Help desk report; Gmail: timestamps in a spreadsheet |
| Time to resolution | Median time from ticket created to closed | Under 24 hours (simple issues); under 3 days overall | Help desk report; Gmail: label closed threads with a date |
| Resolution rate | Tickets closed within 7 days / tickets opened | 90-95% | Help desk; or spreadsheet count |
| Reopen rate | Tickets reopened / tickets closed | Under 8% | Help desk status history; Gmail: threads with replies after closure |
| CSAT | Ratings of 4-5 / total ratings | 85-92% | Help desk survey; Google Form link in your reply |
| Tickets per 100 orders | (Tickets in month / orders in month) x 100 | 5-15 | Orders from Shopify admin; tickets from inbox or help desk |
If you run support from a shared Gmail inbox, a spreadsheet with one row per ticket works. Columns: date received, date first replied, date closed, reopened (yes/no), rating. Five minutes a day keeps it current.
Which popular metrics are noise at low volume?
Net Promoter Score, average handle time, and most agent-level productivity charts are noise at low volume. With 150 tickets a month, a sample of 20 survey responses can swing a score by 15 points on a single answer. Skip them until you pass roughly 1,000 tickets a month.
- NPS: it measures brand loyalty, not support quality. Few support customers answer it, so your sample is tiny and skewed toward extremes.
- Average handle time: it rewards speed over correctness. On small samples, one 40-minute refund investigation distorts the entire month.
- Tickets per agent: with one or two people answering, this is just a headcount division.
- Backlog size: a single snapshot of open tickets. Resolution rate covers it with better context.
The rule is simple. If one customer can move the metric by more than 5 percentage points, it is not a metric yet. It is an anecdote with a chart.
What does 'good' look like by channel?
Good depends on the channel because customers expect different speeds. Chat customers expect answers in minutes, email customers tolerate hours, and social DMs sit in between. A single blended target hides problems, so set first response goals per channel.
Use these working targets for a store at 100-500 tickets a month:
- Email (Gmail, Outlook): first response under 4 business hours; resolution under 24-48 hours.
- Live chat widget: first response under 5 minutes during staffed hours; clearly stated hours outside them.
- Instagram DMs: first response under 1 hour in business hours, since customers treat DMs like texting.
- WhatsApp: first response under 30 minutes in business hours, with the same expectation of conversational pace.
Split tickets per 100 orders by reason, too. If 'where is my order' is 40% of volume, you have a shipping-communication problem, not a support problem. Fixing proactive tracking emails will cut more tickets than any reply template.
How do you build a customer support dashboard for a small team?
A customer support dashboard for a small team needs one spreadsheet tab, six columns of monthly totals, and a trend line. Build it in under 30 minutes, update it monthly, and ignore everything that does not fit on one screen.
A practical setup:
- Log each ticket in a sheet: channel, reason, received, first reply, closed, reopened, rating.
- Create a monthly summary tab with the six metrics as rows and months as columns.
- Pull the monthly order count from Shopify admin for the tickets-per-100-orders ratio.
- Add a median formula for the two time metrics, not an average.
- Highlight any metric that moves more than 20% month over month, then investigate that one only.
A paid help desk automates steps 1 and 2. That is worth paying for once you pass about 300 tickets a month and manual logging eats an hour a week. Until then, the spreadsheet is accurate and free.
Where does Heylio fit in your metrics stack?
Heylio reports two numbers natively: AI resolution rate and handover rate. AI resolution rate is the share of conversations the AI closes on its own. Handover rate is the share passed to a human. The other four metrics you can track in a spreadsheet.
This matters because Heylio bills per AI resolution, so these two numbers are also your cost and workload signals. On the Growth plan at €49 a month, you get 250 AI resolutions. Handing a conversation to a human is free.
Heylio connects to Gmail, Outlook, Instagram DMs, WhatsApp, and an embedded chat widget. It drafts replies from your knowledge base and playbooks, and can look up Shopify order status, cancel orders, and issue refunds.
A rising handover rate is useful, not alarming. It usually points to a gap in your knowledge base. Heylio tracks the corrections your team makes and suggests knowledge base and playbook updates, so the handover rate should fall as those gaps close.
Be realistic about the limits. Heylio does not replace your spreadsheet for CSAT, reopen rate, or tickets per 100 orders. Keep those in your sheet, and compare them before and after you turn the AI on.
What should you do this month?
Start with the six numbers, one spreadsheet, and a monthly 15-minute review. Pull last month's ticket count and order count, calculate tickets per 100 orders, and record your median first response time. That single baseline shows you where to improve first. If AI handles part of your volume, add AI resolution rate and handover rate beside it. Ignore NPS and handle time until your volume justifies them.