Tickets we created from audit findings that haven’t been resolved yet.
At a glance
Live count of HelpScout conversations that Vortex IQ created on your behalf from audit findings (broken checkout flows, refund spikes, performance regressions, ad-on-OOS, missing SEO fields) and which still sit in an unresolved state in the shared inbox. The merchant question this answers is: “how many revenue-protecting fixes have I asked HelpScout to triage that haven’t been picked up yet?” HelpScout calls these conversations, not tickets, but the merchant-facing meaning is the same.
| What it counts | HelpScout conversations created by the Vortex IQ integration (tagged vortex-iq in HelpScout, lowercase-hyphen by HelpScout convention) whose status is active or pending. closed and spam drop out. |
| API endpoint | GET /v2/conversations?status=active,pending&tag=vortex-iq&embed=threads paginated with size=50 (HelpScout default; max 200). HelpScout API v2 is OAuth-2.0 only; the connector runs a refresh-token loop. |
| Conversation-state filter | Active + Pending. Closed + Spam excluded. HelpScout has no separate “On Hold” state, so long-pause conversations stay Active. |
| Bot-handled conversations | Included. HelpScout Beacon’s auto-replies and any custom Workflow auto-responses do not change status by default; the conversation only “closes” when an agent or workflow sets status to closed. |
| Multi-mailbox aggregation | Aggregated across every Mailbox the connector token can read. HelpScout structures multi-brand support as separate Mailboxes (one per support email address). Set vortex_iq.mailbox_filter to scope. |
| Multi-channel scope | All HelpScout channels (Email, Beacon Chat, API, Forms). Vortex IQ-created conversations always come in via API (type: api). |
| Time window | RT, refreshed every 60 seconds. |
| Alert trigger | > 20 open. |
| Anomaly detection | If the count grows by >5 in 60 minutes, Vortex IQ raises an “Audit findings backing up” alert. |
| Time zone | Account timezone in Manage → Company → Time Zone. The HelpScout API returns timestamps in ISO 8601 UTC; the card converts for display. |
| Roles | owner, operations |
Calculation
Calculated automatically from your Help Scout data. See the At a glance summary above for what the metric tracks and the worked example below for a typical reading.Worked example
A US wellness DTC brand on Shopify running HelpScout Plus with three Mailboxes (hello@, wholesale@, returns@). The brand’s CS philosophy is “every customer email gets a human reply”; HelpScout was chosen because the shared-inbox model fits that ethos. Snapshot taken on 02 May 26 at 11:15 ET.
| Mailbox | Open Vortex IQ findings | Notes |
|---|---|---|
hello@ (main consumer) | 9 | Filed during last week’s checkout audit; mix of UX and copy issues |
wholesale@ | 4 | All B2B-tier pricing inconsistencies |
returns@ | 3 | All refund-portal flow issues |
| Total | 16 | Below the 20-ticket alert threshold |
- The mailbox split mirrors how the team already works. HelpScout merchants tend to assign mailboxes to specific people, so 4 wholesale findings and 3 returns findings each have a natural owner. Surface a separate Vortex IQ panel per mailbox in the Nerve Centre; don’t try to drive triage across all 16 in one standup.
- All 16 conversations should appear in the shared-inbox view immediately. HelpScout’s “Inbox” view defaults to showing Active conversations sorted by most-recently-modified. Vortex IQ’s API-created conversations sit in this list with the
vortex-iqtag visible inline. Train the team to triage by tag-color (setvortex-iqto red in Manage → Mailbox → Tags for visibility). - HelpScout’s customer-centric reputation cuts both ways here. The shared-inbox model excels at high-touch consumer support but makes it easy for technical-debt findings to feel “out of place” in the inbox. Pair this card with
shopify.refund_rateso the team sees Vortex IQ findings as protecting the customer experience, not as engineering ticket noise. - The 4 wholesale findings are likely the highest absolute revenue impact. B2B pricing inconsistencies on a wholesale mailbox typically affect 5-figure orders. Cross-reference the wholesale account manager’s open quote pipeline; if any of those quotes touch the affected SKUs, those findings need same-day fixes.
- The card is unsmoothed and counts every Vortex IQ conversation regardless of which Mailbox it landed in. This is the source-of-truth view. For mailbox-level views use the per-mailbox sub-panels.
Sibling cards merchants should reference together
| Card | Why pair it with VortexIQ Findings Open | What the combination tells you |
|---|---|---|
| Abandoned Findings (>14d no movement) | The “silently leaking” subset. | Open count steady but Abandoned rising equals findings filed faster than triaged. |
| Open Conversations (all) | Total HelpScout backlog. | If both elevated together, the inbox is overloaded and Vortex IQ findings get buried. |
| Avg Cycle Time | Triage health. | Cycle time creeping while open count grows means triage is broken. Inspect Top Assignees Overloaded. |
| Finding Resolution Rate (90d) | Throughput counterpart. | Open high + resolution low equals findings filed but never closed, the worst possible audit-programme state. |
| Refund Rate (Shopify / BigCommerce / Adobe) | Refund spikes should generate findings. | Refunds up + findings flat equals audit blind spot; findings up + refunds down equals programme working. |
| Customer Service Sentiment (Shopify) | NPS-side outcome. HelpScout users care intensely about NPS. | Findings dropping + sentiment rising equals proof the audit programme protects retention, the headline narrative for a HelpScout-using brand. |
| Datadog Operational Health Score | Technical findings should map to Datadog signals. | Vortex IQ “checkout error” finding open + Datadog error rate steady equals possible false positive or stale finding. |
Reconciling against the vendor’s own dashboard
Where to look in HelpScout’s own dashboard:Inbox view (secure.helpscout.net/conversations) with the tag filterFor multi-mailbox accounts, use the Mailbox switcher at the top to view per-mailbox; the API aggregates across all mailboxes by default. Why our number may legitimately differ from HelpScout’s view:vortex-iqand the default Active+Pending status filter. Saved Searches, build one fortag:vortex-iq AND status:active OR status:pendingand share with the team. Reports → Conversations filtered by the same tag.
| Reason | Direction | Why |
|---|---|---|
| Time zone | Boundary days off | HelpScout Reports honour the Manage → Company timezone; agent UI honours each agent’s profile timezone. The card uses the company timezone for the alert calculation. |
| Conversation-channel scope | Ours equals theirs | Both include all channels by default. If you scope the saved search to “Email only”, that view is a subset. |
| Archived / deleted conversations | Ours lower | HelpScout’s Trash is excluded from GET /v2/conversations by design. Trashed conversations remain queryable in the UI for 30 days. |
| Pending status semantics | Ours equals theirs | HelpScout’s Pending is “waiting for customer reply”. We count it because for a Vortex IQ finding, “pending” usually means waiting on a developer or agency. HelpScout default views show Pending alongside Active; no semantic gap. |
| Multi-mailbox aggregation | Either | Card aggregates across all Mailboxes; per-Mailbox views are subsets. |
| Webhook lag | Ours lower for “just now” | A conversation just created via the API will be in the Inbox within 1-2 seconds but may take up to 30 seconds to be returned by GET /v2/conversations (eventual consistency). |
| Card | Expected relationship | What causes the divergence |
|---|---|---|
shopify.refund_rate / bigcommerce.refund_rate / adobe_commerce.refund_rate | Refund spikes should generate findings (24-72h lag). | Refunds up + findings flat equals audit blind spot; findings up + refunds down equals programme working. |
shopify.customer_service_sentiment | Inversely correlated. | Open findings climbing past 30 for 2+ weeks predicts CSAT drop of 3-5 points. |
Known limitations / merchant FAQs
Why does Vortex IQ’s number disagree with HelpScout Reports? Three usual culprits:- Time zone. Reports honour the company timezone; agent UI honours each agent’s profile zone. The card uses company timezone for alerts.
- Eventual consistency. The HelpScout Conversations API has 5-30 second eventual consistency on newly-created conversations. If you just filed a finding, give it 30 seconds.
- Trash inclusion. UI views can show Trashed conversations for 30 days; the API excludes them entirely.
status. Add a Workflow that closes Vortex IQ conversations once Beacon confidence threshold is met if you want bot-handled findings to drop out.
We run three Mailboxes (hello@, wholesale@, returns@). Do they all roll up?
By default, yes. The connector token aggregates across every Mailbox it can read. To scope to a single Mailbox, set vortex_iq.mailbox_filter: <mailbox_id> in the connector config. HelpScout merchants typically prefer per-Mailbox panels because their team structure mirrors the mailbox structure.
Open count just jumped, what should I check first?
A 5-step playbook:
- Commerce platform first. Refund Rate, Cancellation Rate, Order Volume on Shopify / BigCommerce / Adobe. Refund spikes drive most audit findings.
- Site reliability. Datadog Operational Health Score, Sentry error rate. A bad deploy generates a tail of customer-facing errors.
- Catalogue. Catalogue Drift, Ads on OOS, Top Products. Pricing or stock changes shipped without QA flood the queue.
- Audit category mix. Filter by Vortex IQ tag (
vortex-iq:checkout,vortex-iq:catalogue,vortex-iq:performance). - Viral negative review or social mention. A bad TikTok or Trustpilot one-star spree drives 30+ tickets in an hour.
vortex-iq to track findings separately from consumer queries.
We use HelpScout for transactional consumer queries AND Vortex IQ findings. Are they double-counted?
No. The card filters strictly by tag. Consumer queries don’t carry the vortex-iq tag and are excluded.
Multi-channel: do Beacon chat conversations count?
Yes if they carry the vortex-iq tag. The card does not filter by type (api/email/chat); only by tag and status.