E-commerce
June 28, 2026
You open Gorgias on Monday morning: 47 open tickets, a CSAT at 76% last week, three "nobody is responding" reviews on Instagram. You don't know if the problem comes from volume, a late agent, a poorly calibrated bot, or a SKU generating returns. Without a consolidated view, you are steering by gut feeling.
Shopify points out that a support dashboard centralizes the essential KPIs needed to make quick decisions: response time, resolution, volume, satisfaction (Shopify, support dashboard 2026). The 2026 e-commerce benchmarks place the median FRT between 2 and 4 hours, the FCR between 65 and 72%, and the CSAT around 80% for a mid-size store (Squire, CX benchmarks 2026).
This guide #216 explains how to build a Shopify support dashboard: metrics, data sources, ops views, and playbooks. Distinct from FCR (#136) and FRT (#137) which drill down into a single KPI each: here is the complete management architecture.
Summary
Why does a support dashboard change Shopify management?
A Shopify support dashboard centralizes into a single interface the indicators that trigger an action: hiring, macro to rewrite, product alert, bot adjustment.
What a good dashboard changes
Monday 9 AM: FRT chat > 30 min? Shift reinforcement or WISMO bot
Wednesday: peak ticket intent return on SKU X? Merchandising alert
End of month: cost per ticket rising? Self-service or AI ROI
Example DTC cosmetics
Shopify store, 2,800 tickets/month, KPIs scattered between Gorgias, Notion, and a spreadsheet. After unified dashboard (3 views section 4): detection in 48 hours of a WISMO spike post-launch palette (obsolete tracking macro), fix, FRT −38%, CSAT +6 pts in 3 weeks without hiring.
Guiding principle
Shopify advises prioritizing an actionable interface rather than 30 decorative metrics. Eight to twelve well-chosen KPIs are better than a wall of numbers that nobody opens.
How does it differ from existing KPI support articles?
Five neighboring pieces of content, five roles. #216 assembles; the others deepen.
FCR (#136)
FCR (#136): definition, benchmarks by intent, macro levers. #216: where to display the FCR in the dashboard and with which filters.
FRT (#137)
FRT (#137): business impact of delay. #216: FRT tile by channel + threshold alerts.
Chatbot KPIs (#11)
Chatbot KPIs (#11): isolated bot metrics. #216: blended dashboard human + bot + self-service.
Support cost
Support cost analysis: financial model. #216: cost per resolution tile updated automatically.
Ticket taxonomy (#135)
Taxonomy (#135): intent tags. #216: "volume by category" dimension of the dashboard.
The #216 promise
3-view architecture, 10 KPIs, Gorgias/Shopify integrations, widget layout, alerts, 30-day playbooks.
How to structure a dashboard into 3 views based on cadence?
Structure the support dashboard into 3 views based on the decision-making cadence, not in a single catch-all screen.
View 1: Daily Ops (5 min, every morning)
Open tickets and backlog > 24h
Median FRT by channel (chat, email, Instagram DM)
Queue: unassigned tickets
Previous day CSAT: alert if < 75%
View 2: Weekly Quality (30 min, Tuesday)
Blended FCR and bot vs. human FCR
Volume by intent: WISMO, return, product, dispute
Reopen rate within 48h
Top 5 macros used vs. top 5 uncovered intents
View 3: Monthly Business (1h, start of month)
Tickets / 1,000 orders (workload ratio)
Blended cost per resolution
Deflection rate self-service + bot
CSAT / returns correlation by product category
Reading Rule
If an Ops KPI slips, fix it this week. If a Business KPI stagnates for 2 months, review policy, help center content, or AI investment. Do not mix timeframes on the same screen.
Which KPIs to include and which formulas to use?
Ten Shopify support dashboard KPIs are enough for a DTC store up to ~10,000 orders/month. 2026 formulas and targets.
Speed KPIs
FRT = first response time − ticket creation. Chat target: < 10 min, email: < 4 h (Squire 2026)
ART (Average Resolution Time) = closure − creation. Target: < 24 h for simple intents
Quality KPIs
FCR = resolved on 1st contact / eligible × 100. Blended target: 75%+
CSAT = satisfied / respondents × 100. E-commerce target: 82%+ (Shopify KPI guide 2025)
Repeat contact = recontact for same intent within 7 days. Target: < 10%
Volume and Load KPIs
Total Ticket volume and by intent (see tag conversations (#117))
Tickets / 1000 orders = normalizes store growth
Efficiency KPIs
Deflection rate = resolved without agent / total. Mature target: 40-60% with bot + help center
Cost per resolution = (support salaries + tools) / closed tickets. Blended AI+human target: €2-3 (Squire 2026)
Tickets / 1000 orders formula
(Tickets month M / Orders month M) × 1000. Example: 900 tickets, 4,500 orders = 200 tickets/1000. Track the trend, not the absolute value: a scaling store may see the ratio drop if self-service improves.
Which data sources should be connected to Shopify?
A useful dashboard connects at least 4 sources on Shopify.
1. Helpdesk (Gorgias, Zendesk, Reamaze)
Primary source: FRT, ART, FCR, CSAT, volume, intent tags, agent performance, bot performance. Gorgias provides native Support Agent and Shopping Assistant reports (Gorgias, AI Agent performance). Weekly CSV export if no API is available.
2. Shopify Admin
Orders, returns, return rate by SKU, actual shipping times. Cross-reference with return intent tickets to detect a problematic product before a wave of negative reviews.
3. Help center / self-service
Help center page views, searches with no results, click-through rate to "contact us" from an article. See measuring self-service. Feeds the deflection tile.
4. AI Agent (Qstomy or other)
Bot conversations, resolved intents, escalations, post-chat bot CSAT. Export or webhook to Looker Studio / Google Sheets.
Minimal DTC Stack
Gorgias native dashboard (Ops view) + Google Looker Studio (Business view) + Notion (weekly qualitative review). No need for enterprise BI at the start.
How do I build the dashboard in Gorgias or Looker Studio?
Two paths to build the dashboard depending on budget and maturity.
Option A: Native Gorgias (2 h setup)
Activate Performance reports → filter by channel and intent tag
Pin FRT, CSAT, volume, and FCR tiles to the team home screen
Create saved views: "Morning Ops", "Weekly Quality", "Individual Agent"
Configure email alerts if chat FRT > 20 min or backlog > 30 tickets
Option B: Looker Studio + exports (1 day setup)
Google Sheets connector powered by weekly Gorgias export (Zapier or cron)
Shopify connector (orders, returns) via app or CSV
Page 1: KPI cards (FRT, FCR, CSAT, volume). Page 2: 90-day curves. Page 3: intent × week heatmap
Global filters: channel, agent, intent tag, period
Recommended widget layout (Ops view)
Row 1: 4 KPI cards (chat FRT, open tickets, 7-day CSAT, backlog > 24 h). Row 2: 14-day volume graph + top 5 intents bar chart. Row 3: agents table (FRT, FCR, individual CSAT). Row 4: list of unresolved escalated bot-to-human tickets.
Color convention
Green = within target, orange = monitor, red = immediate action. Set the thresholds once (section 9) and do not change them every week.
What segmentations should be added to see the real problems?
A flat dashboard hides issues. Add 4 dimensions of segmentation.
By channel
Website chat, email, Instagram DM, WhatsApp, SMS. Instagram FRT is often 3× that of chat: do not average them. See Instagram DM support.
By intent (helpdesk tag)
WISMO, return, product pre-purchase, dispute, invoice. Return intent volume +15% over 2 weeks = product or policy signal, not a general overload.
By SKU / collection
Tag tickets with SKU when relevant. Top 10 SKUs by ticket volume / SKU sales. An SKU representing 8% of sales but 22% of tickets = merchandising priority + dedicated macro.
By agent vs bot vs blended
Compare autonomous bot FCR, agent-only FCR, and post-bot-escalation FCR. If post-escalation FCR < 60%, the bot-to-human handoff is poorly calibrated (lost context).
Example of cross-reading
Chat FRT is OK, CSAT is low, FCR is low for WISMO intent: the bot replies quickly but without a personalized tracking link. Fix the macro, do not hire.
How to configure alerts and review rituals?
Turn the dashboard into an alerting system with pre-set thresholds.
Ops Alerts (immediate notification)
Chat FRT > 30 min during business hours
Backlog > 25 unassigned tickets
Daily CSAT < 70% based on ≥ 5 ratings
Quality Alerts (Tuesday email)
FCR WISMO intent < 85%
Repeat contact > 12% on return intent
Unknown intent volume > 8% (missing tags)
Business Alerts (monthly review)
Tickets / 1,000 orders +20% vs month M-1 without promo
Cost per resolution +15% vs quarter
Deflection −10 pts vs baseline
30-min review ritual
Support manager + ops: go through the Ops view, write down 1 action per red tile, assign owner + deadline. No meeting without a written decision.
Which indicators should be used to measure the impact of the dashboard itself?
Measure the impact of the dashboard on support performance, not just its existence.
Meta-dashboard KPIs
Time-to-detect: delay between volume spike and corrective action
Actions / month resulting from the dashboard (macro, hiring, bot fix)
Team adoption: % of agents opening the Ops view ≥ 4 days/week
A/B process testing
Month 1 without a structured dashboard ritual (baseline). Month 2 with views + alerts + weekly review. Compare FRT, FCR, CSAT. Typical DTC mid-size gain: FRT −25 to 40%, FCR +8 to 15 pts in 8 weeks when alerts trigger macro fixes, not meetings.
Shopify Correlation
Overlay return intent ticket curve vs Shopify Admin return rate. Divergence (tickets go up, returns stable) = communication policy issue. Convergence (both go up) = product issue.
What mistakes should be avoided during implementation?
Five support dashboard anti-patterns that kill adoption.
Error 1: Too many metrics
30 tiles, no one knows what to look at. Fix: 10 KPIs max, 3 section views.
Error 2: Global averages
Average FRT of 2 hours hides chat at 45 minutes and email at 8 hours. Fix: Segment by channel.
Error 3: Dashboard without intent tags
Total volume is useless. Fix: Tag taxonomy #135 mandatory before creating dashboard.
Error 4: Outdated data
Manual export forgotten for 3 weeks. Fix: Zapier or scheduled Gorgias report.
Error 5: KPIs with no owner
Red tile, no one reacts. Fix: Each KPI = 1 owner + threshold + standard action.
How does Qstomy power the Shopify support dashboard?
Qstomy feeds the support dashboard with bot metrics that are directly actionable within your Shopify management.
Exportable Qstomy metrics
Bot conversations / day and autonomous resolution rate by intent
Bot → human escalations with reason (missing data, dispute, explicit request)
Post-conversation CSAT bot vs agent
Pre-purchase product questions converted into orders within 7 days
Uncovered intents: top gaps to enrich RAG
Quantified DTC scenario
In-house brand, 3,400 tickets/month, Gorgias dashboard alone, Qstomy bot untracked. Sync Qstomy metrics → Looker Studio + human+bot "blended FCR" tile. After 8 weeks: bot intent visibility 100%, deflection +19 pts (34% → 53%), blended FRT −42% (4h 10m → 2h 25m), cost per resolution −28%, 3 WISMO macros rewritten following intent FCR alert.
Explore Qstomy analytics, Shopify integration, AI customer support, request a demo.
Which operational playbooks should be launched in 30 days?
Playbook 1: KPI framing (2 h)
List 10 KPIs from section 4. Assign an owner per KPI. Set green/orange/red thresholds. Deliverable: Notion one-pager "Dashboard support v1".
Playbook 2: data connection (4 h)
Activate Gorgias reports. Check intent tags over 90 days (% of tagged tickets > 85%). Export Shopify orders + returns. Connect or schedule Looker Studio refresh.
Playbook 3: 3-view layout (3 h)
Build Ops, Quality, Business views from section 3. Test with 2 agents: do they understand what to do in 5 min? Adjust tiles.
Playbook 4: alerts (1 h)
Configure 6 alerts from section 8. Test notification on a dummy WISMO spike ticket.
Playbook 5: W+4 ritual
Weekly 30-min review on Tuesday + monthly meta KPI review from section 9. Document 3 corrective actions of the month and their measured impact.
Useful links
A Shopify support dashboard is not a BI luxury: it is the control panel of your customer relations. Build it in 30 days, steer it every Tuesday.
Enzo
June 28, 2026
