A support team of four handles around three hundred tickets a week. They are competent, they are fast, and they are permanently behind.
If anybody listed the questions, they would find that roughly twenty of them account for most of what arrives. The same twenty, every week, answered by four skilled people who retype variations of the same paragraphs.
Nobody has made that list. The team is too busy answering the questions to notice that they are the same questions.
The first failure: nobody has counted
The list is short, knowable and stable. A team that tags a fortnight of tickets can see exactly where its time goes, and most never do.
The second failure: the answers live in people's heads
Four agents each with their own version of the same paragraph, none of them written down, all of them slightly different.
The third failure: the help content is written for the company
Organised by product structure and phrased in internal language, so a customer searching in their own words finds nothing.
The fourth failure: the help centre is hard to reach
Buried in a footer, so a customer who would have self served instead writes a ticket, which is the pattern in the cost of answering the same how to question daily.
The fifth failure: the product causes the question
Half of repeated questions are design problems wearing a support costume. A team that reports the top five to the product side changes the volume permanently.
The sixth failure: deflection is done badly
Forcing a customer through articles before allowing contact produces anger. Answering instantly and handing over immediately when it is not a known question produces the opposite.
The seventh failure: the policy questions are treated as repetitive
They are not. An exception request needs judgement, which is the ground in refund and policy exception questions agents repeat.
The eighth failure: the handover is clumsy
A customer who has already explained something and is then asked to explain it again has been given a worse experience than no automation at all.
The ninth failure: the out of hours queue is ignored
Everything arriving overnight waits, which is the problem described in weekend and holiday cover questions.
The tenth failure: new agents learn by copying
Onboarding into a team with no canonical answers means each new agent replicates whatever the person beside them does.
The eleventh failure: nothing is updated after a change
A pricing or policy change should trigger a rewrite of the affected answers, and in most teams it triggers a fortnight of corrections instead.
The twelfth failure: the wording of the answer is never improved
An answer that generates a follow up question is a bad answer, and most teams never notice because the follow up arrives as a separate ticket. Tracking which replies produce a second message is the cheapest quality improvement available to any support function.
The thirteenth failure: the improvement is never measured
Teams that do act on this rarely establish a baseline first, so nobody can say what changed, and the effort is not repeated.
The fix: answer the known twenty instantly and escalate everything else
An assistant that handles the repeating questions in the customer's own words, hands over cleanly with the conversation attached and logs what it could not answer gives four skilled people their week back for the tickets that actually need them. The arithmetic is set out in the hidden hours you lose to repetitive questions. Start at try, see features, and the sector view sits under small business.