Cut first-response time by 12 seconds with a Zendesk macro, then the requester sent two ‘hi?’ nudges that reset the SLA timer and cost us roughly 30 seconds overall. Anyone else ship a tidy optimization only to watch human randomness vaporize the gain inside an hour?
Brutal — the “two ‘hi?’” is why I tag greeting-only end-user updates via a trigger (regex like ^(hi|hello|hey)\b), auto-reply “we’re on it,” and immediately set status back to Pending so Next Reply SLA doesn’t spin up. In Zendesk you can also exclude that tag from the SLA policy: https://support.zendesk.com/hc/en-us/articles/4408886161050-SLA-policies. Would auto-pending mess up your 12s macro stats or is that workable?
Same here — saved about 12s with a macro and then a couple “hi?” pings cost us about 30s. What helped was adding a first line to the macro: “Got it — no need to reply; we’ll update you next,” which cut those nudges by about 40% and stopped our Next reply SLA from bouncing. We skip that phrasing for VIPs who expect a rapid back‑and‑forth.
I started buffering the “hi?” pings upstream: a Gmail filter + tiny Apps Script holds 1–3 word replies for about 90s, sends a single auto‑ack, then releases them as one update so the clock doesn’t yo‑yo; whitelist VIPs/after‑hours to be safe. Building on @darryl_hughes21’s regex, want the script snippet?