Ticket backlog half-life is the median age at which open tickets reach a resolved state, measured across your current queue. If half your open tickets are under 18 hours old and half are older, your half-life is 18 hours. Unlike raw backlog count — which moves up and down with volume and tells you nothing about velocity — half-life isolates one question: is the queue decaying or accumulating? It's the single backlog metric that survives spikes without lying.
Key takeaways
- Ticket backlog half-life is the median resolution age of currently-open tickets, not a count — it measures decay velocity, not pile size.
- Backlog count rises and falls with volume; half-life rises only when your team is losing ground, which is why it survives traffic spikes intact.
- A healthy SMB B2B SaaS half-life sits between 4 and 24 business hours; anything climbing week-over-week is an early warning before SLA breaches start.
- The metric pairs naturally with arrival rate and resolution rate to form a three-number queue model that fits on one dashboard.
- Most helpdesks report median age natively or expose the data via API — you don't need a separate analytics tool to start tracking it.
What ticket backlog half-life actually measures
The term borrows from physics: half-life is the time it takes for half of a population to decay. Applied to a support queue, it's the median age at resolution — half your tickets resolve faster, half slower.
The critical distinction is that half-life is calculated against resolved tickets in a recent window (usually 7 or 14 days), not against the open queue snapshot. You're asking: of the tickets we closed this week, what was the median age at the moment we closed them? That number tells you how fast the queue is genuinely decaying — not how big the pile looks today.
A 12-hour half-life means your team is, on median, resolving tickets within half a business day of their arrival. A 5-day half-life means the typical ticket sits open for nearly a full work week before anyone touches it to completion. Same backlog count, wildly different operational reality.
Why backlog size is a vanity number
Raw open ticket count is the metric every support ops dashboard leads with, and it's almost useless on its own.
Consider two teams, each with 200 open tickets. Team A's tickets are mostly 2–6 hours old because they're slammed with new arrivals but resolving fast. Team B's tickets average 4 days old because the team stopped keeping pace three weeks ago and the count crept up slowly. Backlog count says they're equal. Half-life says Team A is healthy under load and Team B is drowning.
The vanity problem gets worse during traffic spikes. A product launch, an outage, a Black Friday rush — backlog count balloons even on the best-run teams. Leadership panics, agents work overtime, and the metric eventually returns to baseline. But half-life would have told you within 48 hours whether the spike was being absorbed (half-life stable) or compounding (half-life climbing). It separates temporary surge from structural collapse.
How to calculate half-life on real ticket data
The math is straightforward enough that you can run it in a spreadsheet against a CSV export.
- Pull every ticket resolved in the last 7 days (or 14 for less variance on smaller teams).
- For each ticket, compute
resolved_at - created_atin business hours, respecting your team's working schedule and timezone. - Sort the resulting list ascending.
- Take the median value — that's your half-life.
- Re-run weekly and chart the trend.
A few refinements matter in practice. Exclude tickets resolved in under 10 minutes — those are usually duplicates, spam, or auto-resolves that skew the median downward. Segment by channel (email half-life will be longer than chat half-life by an order of magnitude) and by priority, because mixing Urgent and Low into one median hides the signal you actually care about.
Reading the trend: what good and bad half-lives look like
For a 5–15 agent B2B SaaS support team, here's a working benchmark range (these are operational rules of thumb, not industry studies):
| Channel | Healthy half-life | Warning zone | Drowning |
|---|---|---|---|
| Live chat | Under 15 minutes | 15–60 minutes | Over 1 hour |
| Email — standard | 4–8 business hours | 8–24 hours | Over 24 hours |
| Email — complex / engineering escalation | 1–2 business days | 2–4 days | Over 4 days |
| Web form / portal | 6–12 business hours | 12–36 hours | Over 36 hours |
The direction matters more than the absolute number. A team with a 36-hour email half-life that's been stable for six months has a sustainable rhythm — slow, but sustainable. A team with a 12-hour half-life that's been climbing by 2 hours per week for a month has six weeks before it hits crisis. Trend beats snapshot every time.
Pairing half-life with arrival rate and resolution rate
Half-life alone tells you the queue is decaying. The next question is why — and that requires two companion metrics:
- Arrival rate: tickets created per business hour, averaged over the last 7 days.
- Resolution rate: tickets resolved per business hour, same window.
When resolution rate exceeds arrival rate, half-life will trend down. When arrival exceeds resolution, half-life climbs. The gap between them predicts how fast.
These three numbers — arrival rate, resolution rate, half-life — form a complete queue model that fits in three tiles on a dashboard. Backlog count becomes a derived value, not a headline. Most support ops failures come from staring at backlog count instead of this triplet.
How Helptal fits in
Helptal's support ticket reports surface the underlying timestamps you need — firstResponseAt, solvedAt, closedAt — and the response-time report card already exposes median resolution time over 7d / 30d / 90d / 12m windows, which is half-life by another name. For deeper segmentation, the ht_live_* API tokens let you pull the raw ticket data into a spreadsheet or BI tool to compute half-life by channel, priority, or topic. Combined with SLA policies on Growth and Business plans, you get half-life as the leading indicator and SLA breaches as the lagging confirmation.
Frequently asked questions
What is ticket backlog half-life in customer support?
Ticket backlog half-life is the median age at which open tickets get resolved, calculated over a recent window of closed tickets. If your half-life is 12 hours, half the tickets you closed this week were resolved within 12 hours of creation and half took longer. It measures queue decay velocity rather than queue size, which makes it resilient to volume spikes.
How is half-life different from average resolution time?
Average resolution time uses the mean, which is dragged upward by a few very old tickets — one stuck escalation can distort a week of data. Half-life uses the median, so it reflects the typical ticket experience rather than the worst outliers. For backlog health monitoring, the median is more honest because it tells you what's happening to the middle of your queue, not the tails.
What's a good ticket backlog half-life for a SaaS support team?
For a 5–15 agent B2B SaaS team, a healthy email half-life sits between 4 and 8 business hours, and live chat should stay under 15 minutes. More important than the absolute number is the trend: a stable half-life means your team is keeping pace with volume; a climbing half-life means arrivals are outrunning resolutions and you have weeks, not months, before SLA breaches start.
Can I track half-life without buying a separate analytics tool?
Yes. Any helpdesk that exports ticket data with created_at and resolved_at timestamps lets you compute half-life in a spreadsheet in about 10 minutes. Pull the last 7 days of resolved tickets, subtract created from resolved, sort, take the median. Most modern helpdesks also expose median resolution time natively in their response-time reports.
Should I track backlog half-life by channel or in aggregate?
Always segment by channel, and ideally by priority too. Live chat and email operate on completely different time scales — mixing them produces a meaningless blended median. Track half-life per channel as separate trend lines on the same chart, and you'll spot which channel is degrading before the aggregate number tells you something's wrong.
This week, pull a 14-day CSV from your helpdesk, compute median resolution age by channel, and check it against the benchmarks above. Then set a recurring reminder to re-run it every Monday — once you can see the trend, you'll catch queue decay weeks before SLA breaches force you to. If you're evaluating tooling that exposes this data natively, Helptal's free plan includes ticket exports and the response-time reports you need to get started.