Day-of-week x hour heatmap of created conversations — drives staffing.
At a glance
Conversation Volume by Hour is a Support Performance metric tracked from your Intercom workspace. It is your shift-planning map: a day-of-week by hour heatmap of when new conversations are actually created, averaged over 30 days so a single busy day does not distort the pattern. Most response-time problems are really staffing-timing problems — conversations arriving in a window where nobody is rostered. For Blitz, where a sports retailer sees evening and weekend spikes as shoppers browse after work and around fixtures, this card tells you where to put hours so first replies land fast. It carries no sentiment key or alert; it is a planning surface you read, not a metric you threshold.
| What it counts | The count of conversations created in each day-of-week by hour cell over the 30-day window, using each conversation’s created_at. Rendered as a 7 x 24 heatmap. |
| Sample type | API-derived from the Intercom Conversations endpoint, bucketing created_at into day-of-week and hour-of-day cells. |
| Why it matters | Slow response times are usually a timing mismatch, not a headcount shortage. The heatmap shows exactly when demand peaks so you can roster cover where customers actually wait. |
| Reading the value | Darker cells are busier hours. Look for hot bands (evenings, weekends, post-campaign) and check them against your staffing roster — hot cells with no cover are your response-time risk. |
| Currency | number |
| Time window | 30D |
| Alert trigger | — |
| Sentiment key | none |
| Roles | owner, operations |
Calculation
Calculated automatically from your Intercom data. Vortex IQ reads thecreated_at timestamp of every conversation opened in the trailing 30 days, converts it to your reporting time zone, and buckets it into a day-of-week (Mon–Sun) by hour-of-day (00–23) grid. Each cell holds the count of conversations created in that slot across the window; the heatmap shades cells from light (quiet) to dark (busy). There is no threshold on this card — it is built for staffing decisions rather than alerting.
Worked example
A representative reading for Blitz over 30 days. The heatmap lights up in two bands: weekday evenings 18:00–21:00, and Saturday late-morning 10:00–13:00 (right around weekend fixtures and deliveries). Current staffing is weighted to 09:00–17:00 weekdays, which means the two hottest bands fall partly outside cover — and indeed First-Response SLA Attainment dips on exactly those evenings. The action is to shift one agent’s hours later on weekdays and add Saturday-morning cover. After the change, watch the SLA gauge and First-Response Time by Team recover in those windows. Use Vortex Mind to overlay volume with response time per cell, and Ask Viq to answer “what’s our busiest hour on Saturdays?”Sibling cards merchants should reference together
| Card | Why merchants reach for it |
|---|---|
ic_sla_attainment | Confirms whether the hot, uncovered hours are where you miss SLA. |
ic_first_response_by_team | Links peak-arrival hours to which team slows down. |
ic_workload_by_team | Pairs when volume arrives with which team carries it. |
ic_volume_trend | Shows whether overall volume under the heatmap is rising. |
ic_handled_by_agent | Checks whether rostered agents absorb the peaks you staff for. |
Reconciling against the vendor’s own dashboard
Where to look in Intercom’s own dashboard: Open Reports → Conversations and look for a “new conversations by hour” or “volume by time of day” breakdown. Intercom’s own heatmap-style views are the closest equivalent; the underlying timestamp is the conversation’s creation time. Why the Vortex IQ value may legitimately differ:| Reason | Direction | What to do |
|---|---|---|
Time zone. Vortex IQ buckets created_at into your reporting time zone; Intercom may use the workspace or viewer time zone, shifting cells by hours. | Cells shift | Confirm both tools use the same time zone. |
| Created vs first-response time. This card buckets on creation time; a vendor view bucketed on first-reply time will look different. | Variable | Confirm the bucketing timestamp. |
| Window length. Vortex IQ averages over 30 days; a vendor view over 7 days will be noisier. | Variable | Match the window length. |
Known limitations / merchant FAQs
Q: How often does Conversation Volume by Hour update? The card refreshes on the standard data refresh (typically every 30-60 minutes for live integrations). For a real-time read, force a manual refresh from the dashboard. Q: Why does my Intercom report show a different pattern? The usual causes are time zone (which hour a conversation lands in), whether the heatmap buckets on creation vs first-response time, and the window length. Match these before assuming a real divergence. Q: Why creation time rather than when the conversation closed? Because staffing decisions follow arrival — you roster cover for when customers reach out, which iscreated_at. Closing time tells a different story handled by the resolution cards.
Q: Why does this card have no alert?
A volume distribution is neutral. You act on it by comparing it to your roster and to the SLA gauge, which carries the alert. The heatmap’s job is to tell you where to look.