Ask ten merchants how they start the day and you get ten versions of the same thing: open the admin, look at the orders page, scroll, click into a few, glance at analytics, open the inbox, and forty minutes later you have a feeling about the store but no list of what to do. In a June 2026 thread on the Shopify Community about time-wasters, “repetitive checking” of orders and analytics came up alongside shipping setup and app problems as the things that eat the morning.
The routine below replaces the scrolling with four questions asked in a fixed order. Each is a Sidekick prompt you can paste; each takes under three minutes; and once saved as skills, the whole routine is four slash commands. It is what we run on our own store before the first coffee is finished.
The order matters
Money first, then risk, then stock, then yesterday. That order is deliberate. A paid order that has sat unfulfilled for four days is a customer about to email you, so it comes before anything else. A high-risk order that is about to ship is a chargeback you can still prevent. A stockout on a best seller is tomorrow’s lost sales. Yesterday’s numbers are context, not action, so they come last and they are short.
Check 1: what needs a decision today
Give me today’s orders brief, shop timezone: orders since midnight yesterday, paid and unfulfilled, on hold, returns pending, awaiting payment. Then a short table of orders needing a decision: unfulfilled more than 2 days, high risk, address problems, or with a customer note. Then anomalies: any order over 3 times the average order value, a customer with more than 3 orders today, or a product that sold out overnight. Do not change anything.
The five counts tell you the shape of the day in one line. The decision table is the real output: every row is something a person has to look at. On a quiet day it is empty, and that is a fine result. Save it as /orders-today.
What to do with the rows: an order with a customer note usually needs a reply before it ships, so open it. An address problem is a two-minute email that saves a returned parcel. An order over three times your average is either your best customer of the month or a fraud attempt, and the next check tells you which.
Check 2: risky orders before they ship
List unfulfilled orders from the last 14 days with a risk level of medium or high. For each, give the order number, total, risk level, the specific indicators such as mismatched billing and shipping, unusual IP, multiple failed payments or a first-time customer with a large order, and whether the customer has previous successful orders. Suggest one action each: hold and verify, fulfil with signature required, or fulfil normally. Do not cancel or contact anyone.
Shopify’s fraud analysis is strongest with Shopify Payments, where it sees the payment signals directly. The prompt makes Sidekick surface the indicators rather than the label, which is what you need to decide. A first-time customer, a large order and a shipping address in a different country from the billing address is worth a verification email. A returning customer with one mismatched detail is not.
During peak season, run this every morning; the rest of the year, most stores can run it three times a week. If you find yourself acting on it daily, let Shopify Flow hold high-risk orders automatically: the /flow-fraud-hold skill builds that workflow, and the Flow automations post explains it. Save the check itself as /risky-orders.
Check 3: what is about to run out
For every variant that sold in the last 30 days, compute days of cover from current inventory divided by the daily sales rate. List variants with under 14 days of cover, ranked by revenue in the last 30 days, with units on hand and units sold. Separately list variants already at zero that sold in the last 30 days. Do not adjust inventory.
Days of cover beats a flat “alert me under 5 units” because a product that sells one a month and a product that sells ten a day need different thresholds. The ranking by revenue tells you which stockout would hurt most. Most mornings this list does not change, so a glance is enough; when a product jumps into it, that is your restock trigger. Save as /low-stock and, when something needs ordering, follow with /restock, which turns the list into quantities and a supplier email.
Variants with inventory tracking turned off do not appear, which is itself worth fixing: if a product matters, track it.
Check 4: yesterday, in five lines
Give me yesterday’s total sales, order count, average order value and sessions, each compared with the same weekday one week earlier. Then one line with the top 3 products by revenue. Then one line for anything unusual: a refund larger than the average order, a product that sold out, or a traffic spike from one source. Five lines maximum, no headings, no chart.
Five lines, deliberately. The daily number is noisy; what you are watching for is a break in the pattern, and comparing Tuesday with last Tuesday is the fairest way to see one. The weekly review on Monday is where you go deeper; the Monday analytics post covers that. Save as /yesterday-sales.
When a customer is already waiting
If check 1 shows an order with a “where is my order” note, or your inbox has one, answer it before you move on. The prompt that makes this fast pulls the order’s real status:
Draft a reply to a “where is my order” question for order [NUMBER]. Pull the order date, fulfilment status, carrier, tracking number and last scan, and the delivery window promised by the shipping method. Write 3 to 6 sentences stating what is known, the tracking link, and a specific next date or step. If the order is late or has not scanned in 5 days, propose one remedy for me to approve. Do not send anything.
The draft is written from facts, so it never promises a date the carrier has not given. Save as /wismo.
Making it a habit
Four prompts pasted every morning is better than scrolling, but four slash commands is what makes it stick. In Sidekick, type /, choose Create new skill, paste each prompt, and name them with the same first word so they sit together in the menu: orders-today, risky-orders, low-stock, yesterday-sales. If you run a team, share the skills by link so the person on duty runs the same routine you do. The Weekly operations pack includes all of these plus the Monday review, and the planner keeps your 25 slots in order.
Two things the routine will not do. It will not tell you why traffic fell, because Sidekick cannot see your ad platforms; check those separately. And it will not replace looking at the store as a customer once a week. But it does turn the first ten minutes of the day into a list of decisions rather than a feeling, and on most days the list is short.
Questions merchants ask about this routine
I use a 3PL. Does the unfulfilled check still apply? Yes, with one adjustment. Third-party warehouses update fulfilment status on their own schedule, so an order can be shipped but still show as unfulfilled for a few hours. Set the threshold to four days instead of three, and treat the “no obvious cause” group as the list to send to your 3PL contact.
Do I need all four checks every day? Orders and yesterday’s numbers, yes. Risk, daily in November and December and a few times a week otherwise. Stock, daily if you sell fast-moving products, otherwise twice a week is enough because the list rarely changes overnight.
What if I have two stores? Sidekick works per store, so run the routine in each admin. Save the same four skills in both; the shortcuts are yours, the data is each store’s own.
Can I get this without opening the admin? Partly. Turn the checks that matter most into Shopify Flow alerts, such as the low-stock and fraud-hold workflows, and they come to you by email. The Flow post covers both. The orders brief and yesterday’s numbers are still worth a look with your own eyes.
How long until it feels automatic? About two weeks. The first few days you will read every line; after that you will scan for what is different from yesterday, which is what the routine is for.