The pile of open conversations — and why its age tells you more than its size.
Support backlog. A support backlog is the accumulated set of conversations that are open and waiting on the team, beyond what the team clears in a normal working day. A backlog is measured two ways: by count — how many conversations are waiting — and by age — how long the oldest have waited.
A backlog is arithmetic before it is anything else: conversations arrive at one rate and get resolved at another, and whenever inflow exceeds capacity for long enough, the difference piles up. That distinguishes it from a queue — the queue is today’s waiting line, turning over as the team works; the backlog is the debt left when the line stops turning over.
The count is the less informative of the two measurements. Two hundred open conversations from the last two days is a busy week; sixty conversations of which ten are three weeks old is a problem. Age is where the damage lives, so the single most useful backlog statistic is the age of the oldest untouched customer message — a number that cannot be flattered by averages.
Conversations decay as they age. The customer’s question goes stale, the eventual answer needs more context-rebuilding to write, and many customers resolve the wait themselves — by churning quietly. A standing backlog also taxes the team every day: each person scans the same aging pile each morning, feels the same low-grade guilt, and picks the newest, easiest items first, which is precisely how the old ones get older. Backlogs are self-reinforcing; that is why they deserve explicit management rather than hope.
Backlog trend. Weekly backlog trend = conversations opened this week − conversations resolved this week. A positive number means the backlog is growing regardless of how hard the team worked; a flat count with rising oldest-message age means the team is servicing new work while old work rots.
Track three numbers weekly: open count, oldest-message age, and the trend above. Clearing an existing pile takes a decision about order — oldest-first honours the longest waits; newest-first keeps current customers fast while a dedicated rota drains the tail. What doesn’t work is everyone doing a bit of both from one undifferentiated pile in a shared inbox with no assignments. And if part of the pile is genuinely beyond rescue, close it honestly: a short apology naming the dropped ball beats a cheery reply pretending three weeks didn’t pass.
Give your customers faster answers and your team their evenings back.
Desktop and mobile. Free for solo.