Ecommerce virtual assistant tasks fall into four streams: answering customer messages, clearing order exceptions, processing returns, and keeping the catalog clean. A remote ops team works inside your Shopify admin, your Gorgias queue and your Klaviyo account with named logins, so the founder stops being the last line of customer support.
What does a remote ops team actually own?
Split the work by stream, assign an owner, and give each stream a time block. Vague ownership is why queues grow.
| Stream | Daily work | Tools | Owned by |
|---|---|---|---|
| Customer messages | Tickets, chat, social DMs, marketplace messages | Gorgias, Shopify Inbox | Support assistant |
| Order exceptions | Failed payments, address errors, stuck fulfillments, split shipments, carrier issues | Shopify admin, carrier portals | Ops assistant |
| Returns and refunds | Return requests, labels, restocking, refunds, exchanges, disputes | Shopify returns, Gorgias | Support assistant |
| Catalog | New listings, variants, images, alt text, collections, metafields | Shopify products | Catalog assistant |
| Lifecycle email | Flow QA, segment builds, campaign scheduling, post-send reporting | Klaviyo | Ops assistant |
| Reporting | Daily handled log, weekly KPI review | Sheets or Notion | Team lead |
On a small store one person runs all of it part time. Past roughly 1,500 orders a month it usually splits into a support desk and an ops desk. The Wishup 2026 industry report puts administrative work at 31.5 percent of assistant workloads, with admin and marketing together over 62 percent, which matches what e-commerce stores buy: queue clearing plus the recurring marketing operations nobody has time for.
How does the work run inside Shopify?
Grant a staff account with scoped permissions rather than the owner login. Shopify lets you give order management without finances, and app installs without theme editing, so build the role around the job.
What an assistant does in the admin every day:
- Orders. Edit shipping addresses before fulfillment, cancel and reorder, create draft orders for replacements, add order notes and tags that the rest of the team can filter on.
- Fulfillment. Chase orders sitting unfulfilled past your threshold, resolve inventory holds, split shipments, and fix tracking numbers that did not sync.
- Payments. Work through failed and pending payments, flag suspected fraud for review rather than deciding it, and never handle raw card details outside your payment portal.
- Returns. Start returns, issue labels, mark items restocked or damaged, and refund within the ceiling set in the SOP. Shopify’s returns documentation is the reference for the flow.
- Products. Build listings with complete product details: title, description, variants with correct options, SKUs and barcodes, weights for shipping rates, images with alt text, and SEO fields.
- Collections. Keep automated collection rules current, remove sold-out products from featured collections, and check that seasonal collections have been switched over.
- Discounts. Build codes and automatic discounts to a brief, check exclusions and combination rules, and set end dates so nothing runs on forever.
A simple guardrail that saves arguments: write the refund ceiling, the replacement policy and the escalation list into one page, and have the assistant tag every exception with the reason code. The monthly review of those tags is where you find the product or carrier causing the pain.
How do you turn a shared inbox into a queue with Gorgias?
Gorgias works when three things are built: views, rules and macros. Most stores have macros and nothing else.
Views split the queue by what the team should work next: unassigned, awaiting customer, over 12 hours old, returns, VIP customers, negative sentiment. An agent should never have to decide what to pick up.
Rules do the mechanical work: auto-tag by intent, auto-assign by view, auto-respond to order status questions with live tracking, snooze tickets awaiting a carrier update, and escalate anything containing a chargeback keyword.
Macros carry the brand voice. Aim for 25 to 40 covering your real intents, each with variables for name, order number and tracking link, and each reviewed quarterly. Gorgias documentation covers the setup.
Here is a WISMO macro template you can adapt.
Hi {{customer.first_name}},
Thanks for checking in. Your order {{order.name}} shipped on {{order.shipped_date}} and is on its way with {{order.carrier}}.
Tracking: {{order.tracking_url}} Latest scan: {{order.tracking_status}} Expected delivery: {{order.eta}}
Carriers sometimes take a day or two to update scans after collection. If nothing moves by {{order.eta_plus_two}}, reply to this email and I will open a trace with the carrier and sort a replacement or refund.
Anything else I can pick up while I am here?
{{agent.first_name}} at {{store.name}}
Two details make it work. It answers the actual question in the first two lines instead of thanking the customer for their patience. And it sets the next step with a date, which stops the follow-up ticket that would otherwise arrive tomorrow.
What is the plan for cutting WISMO rather than answering it faster?
WISMO is usually the biggest single category, and most of it is a communication problem rather than a shipping problem.
- Send tracking instantly. Trigger the shipping notification when fulfillment is created, not in a nightly batch.
- Publish honest delivery windows. On the product page, at checkout and in the confirmation email. A three to five day promise that holds beats a two day promise that does not.
- Auto-answer order status. A Gorgias rule that recognizes an order number and returns live tracking closes a large share of these before an agent sees them.
- Add a self-service lookup. Link it from the confirmation email, the shipping email and the site footer.
- Work the root causes. Tag every WISMO ticket with a cause: late fulfillment, carrier delay, address error, split shipment, preorder confusion. Review the tags monthly and fix the top one.
- Get ahead of delays. If a carrier region is running late, email affected customers before they email you. One proactive message replaces dozens of tickets.
Where does Klaviyo fit?
Split ownership: the marketer sets strategy and offers, the assistant owns the build and the checks. That division works because the repeatable parts of email are exactly where mistakes happen.
| Klaviyo work | Assistant does | Cadence |
|---|---|---|
| Core flows | QA welcome, browse abandonment, abandoned checkout, post-purchase, winback and back-in-stock flows: links, variables, timing, exclusions | Monthly audit |
| Segments | Build and maintain segments for campaigns, suppress recent purchasers and unengaged profiles | Per campaign |
| Campaigns | Build from approved copy, proof on desktop and mobile, test links, set UTMs, schedule | Per campaign |
| Deliverability hygiene | Manage suppression lists, monitor bounce and spam complaint rates, clean unengaged profiles | Monthly |
| Post-send report | Open, click, placed order and revenue per recipient, with one observation and one recommendation | Per campaign |
| Consent records | Keep sign-up source and timestamp visible on the profile | Ongoing |
The Klaviyo Help Center documents each of these. One rule worth writing into the SOP: nothing sends without a second pair of eyes on the links and the segment. That single check prevents most of the emails a store regrets.
What does a peak season pod look like?
Peak is a staffing and preparation problem, and both have to be solved before the volume arrives.
Six weeks out
- Forecast tickets: last quarter’s tickets per 100 orders multiplied by forecast peak orders, plus a margin for longer handle times.
- Decide coverage: extended single shift, two shifts, or follow-the-sun with a pod and a lead.
- Agree the shift matrix and publish it.
Four weeks out
- Refresh macros for peak: shipping cut-off dates, gift notes, extended returns window, sold-out alternatives.
- Write the promotion brief: codes, exclusions, stacking rules, start and end times in a stated time zone.
- Train the additional assistants on your catalog, not just your tools.
Two weeks out
- Dry run: seed 20 realistic tickets across intents and score the responses.
- Set the escalation tree: who decides on an out-of-policy refund at 11pm, and how they are reached.
- Freeze non-essential catalog changes.
Peak week
- Twice-daily stand-up: backlog, oldest ticket age, top intent, blockers.
- Live dashboard on first response time and backlog, checked at the end of every shift.
- One person owns the carrier exception list all day.
After
- Returns wave staffing for the four weeks following, because returns peak after sales do.
- Retro: what broke, which macro was missing, which promotion generated the most tickets.
Which KPIs prove the desk is working?
| KPI | Definition | Why it matters |
|---|---|---|
| First response time | Median minutes from ticket creation to first human reply | The number customers feel most |
| Tickets per 100 orders | Total tickets divided by orders, times 100 | The health metric for the whole operation, not just support |
| One-touch resolution rate | Share of tickets closed on the first reply | Tests macro quality and agent authority |
| CSAT | Satisfaction score on resolved tickets | Catches fast but unhelpful replies |
| End-of-shift backlog | Open tickets older than your target at shift handover | Prevents a queue quietly compounding overnight |
| Refund and return processing time | Hours from request to resolution | Slow returns generate second and third tickets |
| Listing accuracy | Sample audit of 20 listings a month against your standard | Catches the slow decay nobody notices |
Set your baseline in week one and improve against it. Borrowed benchmarks from other stores are the fastest route to measuring the wrong thing.
A 30-day onboarding checklist
- Shopify staff accounts with scoped permissions, one per assistant, no shared logins.
- Gorgias seats, views, rules and the starting macro set reviewed and approved.
- Klaviyo access with a defined role, and the flow audit scheduled.
- SOP page: refund ceiling, replacement policy, exception reason codes, escalation list with names and hours.
- Brand voice guide: greeting, sign-off, what to apologize for and what not to.
- Carrier portal logins and the trace process for each carrier.
- Catalog standard: title format, description structure, image requirements, alt text rules, SEO fields.
- Reporting template for the daily handled log and the weekly KPI review.
- Baseline numbers captured in week one.
- Weekly 20-minute review booked with the team lead.
AssistBPO staffs e-commerce desks with a named assistant, a backup and a team lead, working your hours inside Shopify, Gorgias and Klaviyo. Assistants are employed, managed staff, never freelancers, and a pod can be scaled for peak and scaled back after. Plans depend on hours, desks and coverage. Get a staffing plan within 1 business day.
Is it cheaper than hiring in house?
The published ranges frame it. The Wishup 2026 cost guide puts managed offshore assistants at $7 to $15 an hour, $640 to $2,400 a month full time, and $350 to $1,300 a month for 20 hours a week. PayScale puts the average US virtual assistant at $19.45 an hour and Indeed at $27.15.
The better question is what the queue is costing now. If the founder answers tickets at 10pm, that is not free, it is the most expensive labor in the business. Move the queue to someone whose job it is, keep the product and merchandising decisions in house, and the store gets both its evenings and its catalog back.
Frequently asked questions
What does an e-commerce virtual assistant do day to day?
They clear the customer queue, then work the exceptions. A typical day starts with the overnight ticket backlog in Gorgias, moves to order exceptions in Shopify such as failed payments, address errors and stuck fulfillments, then returns and refunds, then a block of catalog work: new listings, variant fixes, collection updates, alt text. Weekly tasks sit on top: Klaviyo campaign QA, inventory reorder flags, review responses and a report on the numbers.
Can an assistant issue refunds in Shopify?
Yes, with a staff account scoped to the permissions you choose. Shopify lets you grant order management and refund rights without giving access to finances, payouts or app installs. Most stores set a refund ceiling in the SOP, for example full refunds up to a set order value handled by the assistant and anything above that escalated to the owner. Every refund gets a note on the order stating the reason, which is what makes the monthly review possible.
How do you cut WISMO tickets rather than just answering them faster?
Attack the causes. Send tracking the moment fulfillment is created, not in a nightly batch. Put realistic delivery windows on the product page and in the confirmation email. Add an order lookup page and link it from the confirmation and shipping emails. Set up an auto-response in Gorgias that detects an order number and returns live tracking. Then chase the remaining tickets to their root cause: a carrier, a warehouse, a product or a promotion is usually responsible for most of them.
Who should own Klaviyo, the assistant or the marketer?
Split it. The marketer owns strategy, offers and creative direction. The assistant owns the build and the checks: segment setup, flow QA, link testing, UTM tagging, send-time scheduling, suppression hygiene and the post-send report. That division works because the recurring, checkable parts of email are exactly where errors creep in, and they are also the parts a marketer skips when busy. Keep list growth and consent capture documented, because email consent rules differ by market.
How many assistants does a store need at peak?
Work from ticket volume rather than revenue. Take your tickets per 100 orders from last quarter, apply your forecast order volume for peak, divide by the tickets an agent clears per hour, and add a margin for longer handle times on holiday queries. Then decide coverage: a single extended shift, two shifts, or a follow-the-sun pod. Agree the shift matrix at least four weeks ahead so training lands before the volume does.
What does an e-commerce assistant cost?
Published ranges give you the frame. The Wishup 2026 cost guide puts managed offshore assistants at $7 to $15 an hour and $640 to $2,400 a month full time, with part-time 20 hour plans at $350 to $1,300. PayScale puts the average US virtual assistant at $19.45 an hour and Indeed at $27.15. Plans depend on hours, desks and coverage. Get a staffing plan within 1 business day.
Sources
- Shopify Help Center, returns and refunds
- Shopify Help Center, product details
- Gorgias documentation
- Klaviyo Help Center
- Wishup, The 2026 Virtual Assistant Industry Report
- Wishup, How much does a virtual assistant cost (2026 cost guide)
- PayScale, Virtual Assistant hourly rate, United States, 2026
- Indeed, Virtual assistant salary in United States




