Most store problems are not mysterious. A product is out of stock but still for sale. A payment gateway is failing for one card type. A supplier is three days late. An email flow has been broken since a theme update. Each would have been found within a day if someone had looked, and each cost far more because nobody did.
An operations checklist is the routine that ensures somebody looks. This one is designed for small and growing stores, with owners and evidence for each check, and suggestions for which parts to automate. Treat it as a template to adapt, not a rule. It pairs with the 25-question automation audit and the monthly bookkeeping checklist.
Principles
- One owner per check. “The team” checks nothing.
- Written steps. If you leave for two weeks, someone else can follow the list.
- Evidence. A tick, a note or a screenshot shows the check happened.
- Exceptions first. Look at what is wrong, not at everything.
- Alerts beat memory. Where a machine can notice, let it, and send the result to a channel people read.
Daily
| Check | What to look for | Automate? |
|---|---|---|
| New orders | Count versus normal; flagged, held or failed orders | Alert on unusual count and held orders |
| Order exceptions | Address problems, payment failures, out-of-stock lines | Queue of exceptions with reasons |
| Fulfilment | Orders past the promised handling time | Alert on late orders |
| Support inbox | Oldest unanswered, urgent topics | Triage and tag; see support macros |
| Stock | Items below reorder point, top sellers with low stock | Alerts; see low-stock alerts |
| Site and checkout | Homepage loads, add to cart, test checkout where safe | Uptime and checkout monitoring |
| Payments and disputes | New disputes with deadlines | Alert with due date |
| Ads | Spend versus budget, disapprovals | Platform alerts to one channel |
The daily list should take fifteen minutes when nothing is wrong. If it takes much longer, either the store has too many manual exceptions, which is a signal to automate, or the checks are too detailed for the day. Move detail to the weekly list.
Weekly
| Check | What to look for | Automate? |
|---|---|---|
| Sales and margin | Revenue, orders, average order value, margin, versus last week and last year | Weekly digest; see sales reporting |
| Returns and refunds | Volume, reasons, patterns by product | Report by reason and product |
| Supplier performance | Late shipments, stock-outs, errors | Tracking and exception reports |
| Email and SMS flows | Revenue by flow, deliverability, unsubscribes | Dashboard; spot-check one live message per flow |
| Top and bottom products | What sells, what stalls | Report |
| Customer feedback | Reviews and tickets for repeated issues | Tagging and summary |
| Backups and updates | Platform, theme and plugin updates pending | Update alerts; test before applying on live |
| Automation log | Failed runs, retries, dead letters | Error digest; see webhooks explained |
Monthly
- Reconcile payouts and close the books using the bookkeeping checklist.
- Count inventory for high-value items and compare with the system.
- Review ad performance by campaign, with return on profit, not just revenue.
- Review every automation: is it running, who owns it, did it fail, is it still needed?
- Check sales tax thresholds and registrations.
- Read twenty random tickets and ten automated emails as a customer.
- Review shipping costs against carrier invoices.
- Update the KPI dashboard and note actions.
Quarterly
- Policies. Shipping, returns, privacy and terms. Do they match what you really do?
- Suppliers. Review prices, reliability and alternatives.
- Access. Who has logins to your store, ads, email platform and automations? Remove those who no longer need them.
- Credentials and keys. Rotate where appropriate and remove unused apps.
- Marketplace policies. Check for changes on every channel you use.
- Disaster plan. Where are the backups, and what happens if the store goes down?
- Audit. Repeat the automation audit.
Seasonal and event checks
Before a major sale, run a short extra list: test checkout on mobile, raise inventory buffers, check that the site can handle the traffic, confirm carrier cut-off dates, brief support, prepare holding messages for delays and set alerts at a lower threshold. After the event, review what broke. See order automation rules for guard rails worth reviewing in advance.
Who does what
| Area | Typical owner in a small team | Backup |
|---|---|---|
| Orders and fulfilment | Operations lead | Owner |
| Support | Support lead | Operations lead |
| Stock and suppliers | Buyer or operations | Owner |
| Marketing and flows | Marketing lead | Owner |
| Money and reporting | Bookkeeper or owner | Accountant |
| Automations and site | Technical lead or agency | Named alternate |
In a very small store, one person fills most of these roles. The important thing is that the list exists and that a second person knows where it is.
Keep a short log
A simple shared document or table, with date, check, result and action, is enough. Its value shows up when a problem recurs: the log tells you when it started, what changed and who dealt with it. Add a column for “automated or manual” so you can see which checks still depend on someone remembering.
Write procedures, not just checks
A check tells you what to look at; a procedure tells you what to do when it fails. For each recurring exception, write three lines: how to recognise it, what to do first and who to tell. “Order held for fraud review” might read: open the order, compare the billing and shipping addresses and the customer history, release or cancel within four working hours, and note the reason. Procedures turn a five-minute puzzle into a two-minute task, and they are the first thing a new team member or an agency needs. Keep them in one place and date each one so you know when it was last reviewed.
| Exception | First action | Escalate when |
|---|---|---|
| Payment failure spike | Check gateway status and recent changes | More than a few failures in an hour |
| Supplier delay | Check tracking and contact the supplier | Delay passes the promised date |
| Stock mismatch | Compare the system with a physical count | Difference above your threshold |
| Angry customer | Respond within the target time with a clear next step | Money, safety or legal issue |
| Broken automation | Read the error, retry once if safe | Second failure or any financial effect |
A sample morning digest
A useful digest is short and sorted by what needs action. For example: orders in the last 24 hours compared with the usual range; orders held or failed with reasons; orders past handling time; products below reorder point; open disputes with deadlines; support tickets older than a day; ad accounts with disapprovals; and the result of the checkout test. Anything normal is a single line. Anything abnormal links straight to the record. The test of a good digest is that on a quiet day you can read it in under a minute and trust the silence.
Automate the checking
Checklists are most reliable when software performs the collection and a person reads the result. A morning message that says “48 orders, 2 held for review, 1 late, 3 products low, checkout test passed” takes seconds to read and replaces several logins. Our ecommerce workflow automation and care plans build and watch these digests, and customer support and order care covers the support side.
Making the checklist stick
A checklist only works if it is used. Put it where the work happens: in your task tool, as recurring tasks with an owner and a due time, rather than in a document that nobody opens. Keep the daily list short enough to finish in a quarter of an hour. Review it in the weekly meeting for two minutes: what was missed, what found a problem and what could be automated. When someone finds an incident that the checklist did not catch, add a check for it the same week. Once a quarter, delete anything that has never found a problem and does not protect against a serious one. A living checklist of twenty useful checks beats a perfect one of a hundred that everyone skips.
A worked example
A store with three staff takes about 90 minutes each morning to check orders, stock, support and ads across five logins. A digest and alert setup cuts it to 20 minutes of reading and acting on exceptions. That saves about 70 minutes a day, or around 25 hours a month. At $22 an hour, that is $550 a month. More important, an alert on a failed payment gateway catches a problem within minutes instead of at the next morning’s check; if the gateway failure would have cost 15 lost orders at $65, the alert alone saves about $975 once. The figures are illustrative.
Mistakes to avoid
- A checklist nobody owns.
- Too long a daily list, so it is skipped.
- Checking without recording, so you cannot show it happened.
- Alert overload, so people mute the channel.
- Never pruning checks that no longer matter.
- Testing only on desktop. Most shoppers use phones.
Sources and further reading
Figures and rules in this guide come from the primary pages below. Platform limits and prices change, so open the page before you build on a number.
- Shopify Help Center: orders, managing and fulfilling orders
- Google Search Central: site status, monitoring crawl and indexing issues
- Stripe Docs: disputes, dispute response deadlines