100% secure — encrypted, never shared.
All case studies
Home & electronics

A two-store seller puts eBay seller automation on replies, listings, and payouts

Buyer questions, listing drafts, and payout checks used to split one owner's week. One workspace now covers messages, listings, and the money.

Marketplace
eBay
Stores
2
Team
Owner + 1 assistant
Messages
~90 a day
  • <30sMedian first reply
  • 2×Listings drafted each week
  • 1 viewPayout status across both stores

The challenge

The owner packed by day and answered buyers at night across two stores — one for open-box home electronics, one for used phones and small appliances. Kettles, stick vacuums, and sealed small appliances lived on the first store. The second store was phones, earbuds, and the odd tablet, each one photographed on a kitchen table after the school run. A message at 11 p.m. about a missing charger usually waited until the next packing shift. Buyers on the west coast, and buyers overseas, were awake while the bench was dark. A few of them opened a case before anyone replied, so the morning went to a dispute instead of to orders that still had to ship.

The assistant arrived after lunch, printed labels, and worked through the pile until the carrier cutoff. Anything that was not in a box by then rolled to the next day, and so did the inbox. The owner tried to clear messages between pickups: a tracking question in the car park, a cable question at the kitchen counter, a return request after the children were in bed. None of those replies were hard. Each one still required the order, the listing, and a sentence that did not contradict either. On a ninety-message day, that was the whole evening.

Off-the-shelf eBay automation tools pasted one template on every thread, and an eBay chatbot could not see a tracking number. Those automated eBay messages created a second wave of “that is not my order” replies. The owner turned them off after a week. A wrong carrier, or a cable the photos never showed, took longer to unwind than typing the reply. One buyer was told a vacuum had shipped with a spare filter that belonged to a different model. Another was given a delivery date from an order placed the week before. The assistant stopped trusting anything sent without a look at the order, and the owner went back to typing, which at least was slow in a way they understood.

New stock sat in photos on the kitchen table while titles were typed by hand. Six or seven listings was a good evening, and a sourcing day waited until the weekend. eBay sales optimization never got past the listings they had time to finish. Titles stayed short, item specifics were skipped, and condition notes were copied from the last similar item, including storage size and battery health on the phone store. A 128 GB phone and a 256 GB phone shared a title more than once. The correction happened after the sale, in a message, which is the expensive way to write a listing. Each return ate the margin on the next two sales.

About half the messages were the same three questions: where is my order, can you ship today, and does this include the cable or the box. Each one still meant opening the order and writing a reply that matched the listing. The assistant was packing during that window, so routine questions buried a wrong color or a phone that would not power on. Those were the messages that actually needed a person, and they sat under a pile of tracking questions until the buyer had already asked twice. By the second ask, the tone was worse, and the reply took longer because it had to apologize for the delay before it answered the question.

The two stores made the pile harder, not just larger. A buyer on the appliance store asking about a voltage plug did not belong in the same mental list as a buyer on the phone store asking about Face ID. The owner kept both logins and switched between them. The assistant knew the appliance store well and the phone store less so, so phone questions waited for the owner even when the answer was on the listing. Shared passwords were the workaround, and anything unsure waited until night.

Payout questions lived in a spreadsheet rebuilt on Sundays: order, buyer, and whether the funds had been released. Holds and failed payouts were easy to mix up across two stores, because the sheet was a memory of Seller Hub rather than a live view. A hold on the phone store once sat unnoticed for nine days. The owner found it while searching for a different order a buyer had asked about. The funds were not missing. They were sitting in a status the sheet had copied wrong the previous Sunday, and nobody had opened that row since.

How ZygenAI helped

  • Buyer messages

    ZygenAI is the shop's eBay message automation. Before a reply goes out, the agent reads that buyer's order: tracking number, carrier, delivery date, and whether the item has shipped. eBay buyer message automation and AI customer service for eBay sellers reply from those facts, in a short tone the owner set once. eBay customer service automation covers combined postage, handling time, and whether the parcel has left. With no tracking yet, the reply says so and quotes the handling window instead of inventing a scan. Refunds, “not as described” claims, and angry messages stay with the assistant. That is the eBay customer support software they run day to day: a short list after packing, not a ninety-message inbox. The owner set the tone in one sitting: short sentences, the carrier named only when the order has one, and no delivery day the order does not show. The first week they read every sent reply. Most were the tracking answers they used to type at night.

  • What the afternoon looks like

    The assistant still prints labels after lunch and still hits the carrier cutoff. The difference is the screen next to the labels. Escalations are the threads a person has to touch: a buyer who says the kettle arrived dented, a phone that will not charge, a request to cancel after the label is already on the box. Tracking questions are no longer in that pile. If a buyer writes again because the first reply was wrong, that thread comes back to the list, which is how the owner notices a bad instruction. On a normal day the list finishes before the school pickup. On a late-courier day it waits until evening, and it is still exceptions rather than the whole inbox. The owner checks it once, and stays only if a refund needs a decision.

  • Rules for each store

    Instructions differ by store, because a sealed kettle and a used phone cannot share a reply. The open-box store never promises an accessory the photos do not show. The phone store names storage only when the listing already does, and sends battery questions to the condition note. A buyer who asks whether a phone is unlocked gets the listing's answer, not a guess from the last phone that sold. Anything else, including an offer under the listed price, goes to the assistant with the thread intact. The owner still samples sent replies. One early instruction offered combined postage on any two items. That was true for small appliances and false for a phone plus a vacuum, so the line was rewritten to name the appliance store only.

  • Listings

    Listing Studio is their eBay selling software. Photos from the table — front, ports, wear, and what is included — become a draft instead of a blank title. These eBay seller tools and eBay selling tools produce a title, item specifics, and description to review before anything goes live. Someone still checks the model number against the device in hand, because a wrong storage size becomes a return. A draft that mentions a charger is deleted unless the charger is in the photos. With the inbox covered, the same AI tools for eBay selling and eBay seller productivity tools let one assistant work both stores, scoped to those stores only, without sharing the eBay password. They can open the phone store and the appliance store from the same workspace, and they never see the owner's login. A second person added for a busy week is invited to one store and removed when the week ends.

  • The listing routine

    A sourcing morning ends with drafts waiting. The owner fixes condition language, deletes a detail the pictures do not support, and publishes when the price looks right. The phone store moved off a weekend backlog, and the extra listings did not bring the old inbox back. Photograph, draft, check the model against the device, list. That fits between the school run and the carrier cutoff on days the owner is home. On sourcing days the photos wait until evening, but the evening is no longer also the inbox. The assistant publishes appliance drafts the owner has already priced. Phone drafts stay with the owner, because a wrong storage figure is the mistake they refuse to delegate. The owner still changes a price before anything goes live.

  • Payouts

    eBay payout management sits next to the orders for both stores, so Sunday no longer starts by copying payout status into a sheet. eBay payout tracking, eBay seller payment management, and eBay seller financial management show which orders are paid, held, or failed. A held order is on the same screen as the buyer and the item. The assistant can see that a refund is waiting on a hold without opening a spreadsheet that might be a week old. The same eBay sales management software is the weekly review: what shipped, what is still in handling time, and which payout needs a look before the next sourcing run. The owner still notes what a sourcing run cost. Whether the marketplace has released the funds is no longer reconstructed from memory on Sunday night.

The result

The shop now runs on eBay seller automation: eBay automation software for replies, automation tools for eBay sellers for the questions that repeat, and AI tools for eBay sellers that only speak from order facts. First reply sits under thirty seconds, including after the packing bench is shut. A buyer who writes at 11 p.m. about a charger gets the order's answer then, not the next afternoon. Refunds, photo mismatches, and hold requests still reach the assistant, and that list clears after dispatch. The owner did not stop reading the shop. They stopped reading the same tracking paragraph ninety times.

The owner checks escalations and payout exceptions once a day. The assistant no longer copies tracking numbers or rebuilds a payout sheet. Eleven hours a week moved from the inbox and the spreadsheet to sourcing and photographs, and listing drafts doubled. Those are the evenings that used to start with the inbox and end with six listings. They now start with the photos. A phone that comes home on Thursday is listed on Thursday, after the storage is checked against the device.

The stores stay separate. Phone replies do not borrow open-box condition language, and a payout hold on one account is not buried under the other. The owner still samples replies and edits an instruction when a follow-up shows a gap. They have not handed the shops over. The repetitive questions are answered from the order, the listings start from the photos, and the money is visible without a second copy of it.

What did not change is the part that makes the margin. The owner still chooses what to buy, still sets the price, and still decides when a return is worth fighting. The assistant still packs and still writes the replies that involve an apology. Buyers who want a discount still wait for a person. The messages that can be answered from the order no longer compete with the ones that cannot.

Start Now

Illustrative case studies based on typical seller workflows. Individual results vary with store size, category, and settings.

Stop typing the same replies.

Connect your store. Let AI handle customer support.

Start NowFrom $14.99/month. Cancel anytime.