The SmallOffice Brief Subscribe
Customer Service
Sat, Nov 14, 2026
GPT-4

82% of queries, $106K saved yearly

An Australian online home-goods retailer takes in $6.8M a year. It holds 2,800 products and gets more than 1,400 support tickets a month. A 4-person team worked a shared Gmail inbox, weekdays only. Weekend tickets waited 28 hours or more. The vendor indexed 340 help articles and all 2,800 products, and connected live stock, pricing and shipment tracking. The assistant answers from that data only. It starts returns for standard cases and makes the label. Frustrated or complex cases go to the team with the full conversation. The build ran 3 weeks in shadow mode inside a 16-week project.

How it wentQueries resolved with no person rose from 0% to 82%. First response fell from 11 hours on weekdays and 28 hours on weekends to under 8 seconds. Human tickets fell from more than 1,400 a month to about 250. Satisfaction rose from 71% to 94%. Support cost per ticket fell from $11.20 to $3.40. The vendor claims $106K a year in savings. The retailer is not named and no outside audit is cited.
Try it · based on the source's workflow
Answer customer questions for [store] from [help articles], live [Shopify] stock and pricing and [ShipStation] tracking only. Start returns for standard cases and generate the label. If the customer turns frustrated or the case is complex, escalate to [team] with the full conversation. Show me every escalated thread the same day.
A starting template based on the workflow the source describes. Review it before you use it; you are responsible for what it does.
Source: I-Verve case study · Topic: Customer Service · Tool: GPT-4
ShareXLinkedInForward to a friend
Cite this case82% of queries, $106K saved yearly. The SmallOffice Brief, Sat, Nov 14, 2026. https://smalloffice.space/case-82-of-queries-106k-saved-yearly
Share this win
82% of queries, $106K saved yearly — from The SmallOffice Brief, a daily brief of real AI cases for small businesses: https://smalloffice.space/case-82-of-queries-106k-saved-yearly
One bot, one job

Start with one job only. Fill in the three lines. Keep the bot on that job.

The one job this bot does: ______________________________

What it may do on its own: ______________________________

What it must ask me first: ______________________________

Before and after

Two lines from this case, side by side. No new figures.

Before · as the case describes it

An Australian online home-goods retailer takes in $6.8M a year. It holds 2,800 products and gets more than 1,400 support tickets a month. A 4-person team worked a shared Gmail inbox, weekdays only. Weekend tickets waited 28 hours or more. The vendor indexed 340 help articles and all 2,800 products, and connected live stock, pricing and shipment tracking. The assistant answers from that data only. It starts returns for standard cases and makes the label. Frustrated or complex cases go to the team with the full conversation. The build ran 3 weeks in shadow mode inside a 16-week project.

After · as the source reports it

Queries resolved with no person rose from 0% to 82%. First response fell from 11 hours on weekdays and 28 hours on weekends to under 8 seconds. Human tickets fell from more than 1,400 a month to about 250. Satisfaction rose from 71% to 94%. Support cost per ticket fell from $11.20 to $3.40. The vendor claims $106K a year in savings. The retailer is not named and no outside audit is cited.

The after line is the source's claim, not our measurement.

smalloffice.space · Get a case like this in your inbox each morning. Subscribe free

Send this case to my inbox

Read this case on your phone now, and keep it for later. Type your email. We send this case to that address, and you get the next cases each morning too.

Additional Details

Published: Sat, Nov 14, 2026 · Updated: Sat, Nov 14, 2026 · Topic: Customer Service · Tool: GPT-4 · Source: I-Verve case study
Next case: 47% of 700 tickets, resolved alone. Get cases like this in your inbox each morning.
5 cases in your inbox each morningSubscribe free