Shopify Weekly Numbers Review: A 30-Minute Routine

Quick answer: Spend 30 minutes on the same weekday every week, in a fixed order. First, check last week’s orders and net sales against the previous four weeks in the Shopify Analytics dashboard (5 minutes). Second, work out which of sessions, conversion rate or average order value moved and look only inside that branch, opening Shopify reports first and GA4, Search Console or Klaviyo only if the branch calls for them (20 minutes). Third, choose one change and write it down with a date to check it (5 minutes), and leave the rest of a long KPI list for the monthly review.

Search for what to review on a Shopify store each week and you get lists of KPIs. They do not tell a busy founder what to open first, what to ignore this week, or when to stop. A list of metrics is not a routine.

This post is the routine: a fixed order, a time budget for each step, and the Shopify, GA4, Search Console or Klaviyo report to open at each one. It is written for the owner of a Shopify store doing roughly A$150K to A$1M a year, who looks at the numbers themselves.

It is the companion to the three numbers Shopify founders watch: that post says which metrics deserve a permanent place on your wall, and this one says what to do on Tuesday morning with the screen in front of you.

One note on scope. Prism3 does SEO, GEO and CRO for Shopify stores. We do not run ads or email, but a weekly review that ignores those numbers is incomplete, so the routine says where to look. It also helps you tell, from your own numbers, whether the problem is traffic, conversion or order value before you spend time or money on a fix.

Report names below are as they appear in Shopify’s, Google’s and Klaviyo’s help centres as at 31 October 2026. If a menu label has moved, search the platform’s help for the report name.

The routine at a glance

The order matters more than the exact minutes. Orders and revenue come first because they are the outcome. Everything else is a possible explanation, and starting with explanations is how a review turns into an hour of browsing.

Before you start: the five-minute setup

You do this once, then it takes a minute each week.

Pick one day and one reporting week

Choose a fixed weekday and review the last full week, Monday to Sunday. A review done “when I get a chance” compares unlike periods.

If you will open GA4 at all, make it Tuesday or Wednesday. Google’s Analytics Help says data processing can take 24 to 48 hours, and that report data may change during that time, so Sunday’s numbers read on Monday morning may not be final. Search Console’s help says its newest data can be preliminary too. Shopify’s overview dashboard is current within about a minute of opening, per its help, so Step 1 is never held up.

Set up the log sheet

One spreadsheet, one row per week. Three groups of columns is enough:

If you run ads, add total ad spend for the week to the outcome column. ROAS, MER and blended CAC explains why store revenue against total spend is safer to log than any single platform’s own figure.

Compare against the last four weeks, not only last week

Last week alone is a weak comparison, because one odd week makes the next one look odd. Log the last four weeks and compare this week with that range, lowest to highest. The Shopify dashboard’s comparison setting offers Previous period, Previous year, a custom range or none (Shopify’s help). Use Previous period as your default, and your log for the four-week range.

Two exceptions:

• If the previous four weeks included a sale (Black Friday, a Boxing Day sale, a launch), they are not a fair baseline. For orders and net sales, use Previous year or the same weeks in your log from last year. If you are reading this in December, that probably includes you.
• Do not use Previous year for sessions or conversion rate. The reason is the Shopify change below.

Know about Shopify’s September 2026 session change

This matters for any store reading conversion rate right now.

Shopify changed how Analytics counts sessions, with a rollout from 21 to 23 September 2026 (Shopify Help Center, session measurement update; Shopify changelog, 21 September 2026). Per Shopify, a session now ends after 30 minutes of inactivity instead of at midnight UTC; some sessions without a pageview are now counted, such as a customer going straight to checkout from a cart link; and identified bot sessions are filtered out of session reports by default (the Human or bot session filter lets you include them, and Shopify says the Home page and Live View always filter bots).

Shopify says orders, sales and customer counts are not affected, and the changelog lists the affected session-based metrics as sessions, conversion rate, add-to-cart rate, reached checkout rate, checkout conversion rate, bounce rate, pageviews per session, online store visitors and searches. Because conversion rate is calculated from sessions, it can shift even when orders and sales do not: Shopify says that if sessions fall because bots are filtered out the rate might increase, and if sessions rise because no-pageview sessions are counted it might decrease. Shopify tells merchants to treat the update as a new baseline, and notes that using the same bot-filter setting does not make sessions before and after the update directly comparable. Bot classification also only applies to sessions from 7 October 2025 onward.

What this means for your weekly review:

1. Conversion rate and session counts from before about 21 September 2026 are on a different yardstick. Do not compare this week’s conversion rate with the same week last year, or with August.
2. Start your conversion baseline from after the change. By mid-December you have a clean run of weeks.
3. Orders and net sales are still comparable across the boundary, which is why Step 1 uses them.
4. Use the same Human or bot filter setting every week wherever the report offers it.

This is a dated statement. If Shopify changes session counting again, its changelog and help page are where to check.

Step 1 (minutes 0 to 5): orders and revenue

Open Analytics in your Shopify admin. The overview dashboard’s default range is the last 90 days, per Shopify’s help, so change it to last week (Monday to Sunday) and set the comparison to Previous period. Click a card’s title and Shopify opens the report behind it.

Open two reports from Analytics > Reports:

• Total sales over time shows the number of orders and total sales over time, grouped by a unit of time you choose (Shopify Help Center, sales reports).
• Orders over time shows the number of orders, average units per order and average order value (Shopify Help Center, order reports).

Write four numbers into your log: orders, sales, average order value and sessions. You will not interpret them yet. You are just recording them.

Pick one sales measure and never switch

Shopify defines the measures differently, and the difference matters:

• Gross sales is product price times quantity, before taxes, shipping, discounts and sales reversals.
• Net sales is gross sales minus discounts minus sales reversals.
• Total sales is gross sales minus discounts minus sales reversals, plus taxes, duties, shipping charges and fees.

(Definitions from Shopify Help Center, sales reports.)

I prefer net sales as the weekly revenue number, because shipping, taxes and fees do not measure demand for your products. Shopify’s help does not list net sales as a column of the Total sales over time report, so check which sales measures your Analytics reports offer (you can customise a report as a new data exploration). If you can only get total sales, use that, every week. What is not fine is switching between them, because a change in shipping charges alone can move total sales without a single extra customer.

The decision rule: is this week outside the usual range?

Weekly order counts wobble even when nothing has changed. For a small count, the wobble is large in proportion. As a rough rule of thumb (statistical arithmetic for random counts, not a Shopify figure), the typical swing either side of the average in a count of random events is about the square root of the count:

So a store with 40 orders a week should not react to a 10% dip. That gives you a working rule:

• Inside last four weeks’ range, and within the typical swing: nothing moved. Do not invent a story. Go to the “quiet week” routine below.
• Outside the four-week range, or past the typical swing: something moved. Go to Step 2.
• Moved in the same direction two weeks running: treat it as worth a closer look even if each week alone is inside the swing.

A good week inside the range is not evidence that last month’s change worked.

End Step 1 with one sentence in the log, for example: “Orders down 20% on the four-week average, net sales down 19%.” If you cannot write that in five minutes, you have opened too many tabs.

Step 2a (minutes 5 to 10): find the branch that moved

Revenue is, roughly, orders times order value, and orders are sessions times conversion rate. So net sales can fall for three reasons:

1. Fewer people arrived (sessions fell).
2. The same number arrived and fewer bought (conversion rate fell).
3. The same number bought, but each spent less (average order value fell).

Look at the three drivers in your log against the four-week range and decide which one is outside it.

Open Conversion rate over time in Shopify to get the conversion figure for your log. Shopify defines it as the percentage of sessions that resulted in a purchase (Shopify Help Center, behaviour reports), and it can be grouped by day, week or month. Remember the September 2026 note above: judge it only against weeks since the change.

If two drivers moved, take the one with the bigger effect on revenue. If you cannot tell, take conversion, because it is the branch you control most directly. Why a Shopify conversion rate drops covers the usual suspects.

A worked example (illustrative numbers)

Here is one week for an invented store. These are made-up figures to show the arithmetic, not a benchmark and not a client result.

Orders are down 20% and net sales are down about 19%. Sessions are inside their range and AOV is flat, so the branch is conversion: the same number of people arrived and fewer bought. Fourteen fewer orders is also larger than the roughly eight-order swing you would expect from a 70-order week, and below the lowest of the last four weeks (66), so this is worth acting on. We will come back to this store in Step 3.

The quiet week: when nothing moved

Most weeks, nothing moves outside the range. Use the spare minutes for a standing check, rotating through four:

• Week 1: place a test order on your phone, on mobile data, using each payment method you offer (refund or cancel it afterwards), and note any step that feels slow or confusing.
• Week 2: compare orders in GA4 with orders in Shopify for the week. GA4 vs Shopify Analytics explains why they will not match exactly and what gap is normal.
• Week 3: open Search Console and look at non-brand clicks and the pages that lost clicks (steps under the traffic branch below).
• Week 4: open the Klaviyo dashboard and check that each flow you rely on is still live and sending.

Step 2b (minutes 10 to 25): look inside the branch that moved

Pick the one section below that matches your branch. Ignore the other two this week.

If traffic moved

You are answering one question: which channel changed? Work from the broadest view to the narrowest.

1. Shopify: Sessions by referrer. Shopify’s help describes this as showing whether visitors come directly, from search or from a referral (Acquisition reports). It tells you the shape of the change.

2. GA4: Traffic acquisition. In the left menu, go to Reports > Acquisition > Traffic acquisition. Google’s help says it shows where new and returning users come from, with session default channel grouping as the default dimension, and columns for sessions, engaged sessions, engagement rate, key events and total revenue. Google defines an engaged session as one lasting 10 seconds or longer, with a key event, or with two or more page or screen views.

Compare the channel rows for the two weeks. A channel whose sessions and engagement rate both fell is a different problem from one whose sessions simply stopped arriving.

Google’s help says the User acquisition report is scoped to new users and Traffic acquisition to new sessions, so for what is bringing people to the store this week, use Traffic acquisition. Use GA4 to find the channel and Shopify for the revenue figure.

3. Search Console, if organic search is the channel. Open Performance > Search results. Google’s help describes four measures: clicks, impressions, click-through rate (CTR) and average position.

• Set the granularity to Weekly. Google’s help says weekly and monthly views smooth out daily fluctuations such as weekends and holidays.
• Use the Date filter and choose Compare to set last week against the previous week. Google’s help recommends weekly or monthly granularity for comparisons, and allows only one comparison at a time.
• Open the Queries and Pages tabs and sort by the Difference column, which the help describes as the way to find queries or pages with significant change.
• If your property has it, use the branded queries filter to split brand from non-brand clicks. Google announced it on 20 November 2025 and said on 11 March 2026 that it was available to all eligible sites. It works only on top-level properties with a sufficient volume of queries and impressions, and Google says some queries may occasionally be misidentified. If you cannot see it, filtering queries by hand for your brand name is the fallback.

What to conclude: if brand clicks fell, look at what you did (or stopped doing) in marketing a few weeks ago. If non-brand clicks fell on a few pages, look at those pages. If impressions fell but CTR held, you are showing up in fewer searches; if impressions held and CTR fell, something else on the results page is taking the click. The organic traffic drop diagnosis walks through that sequence.

4. Klaviyo, if email is the channel. The Klaviyo Home dashboard shows total and attributed revenue by channel and by source (flows or campaigns), with a comparison to a previous period, plus top-performing flows and recent campaign messages with open and click rates (Klaviyo’s help). Check whether a campaign went out last week, and whether each flow is still live. The figures follow the conversion metric and time period you select on the page. Attributed revenue is Klaviyo’s own attribution; email attribution across Klaviyo, GA4 and Shopify explains how it relates to the other two.

5. Paid, if you run ads. Open the ad platform and look at spend and delivery for the same two weeks: did a campaign pause, hit a budget cap or get rejected? We do not run ads, so I will not advise on platform detail.

Decision for the traffic branch: name the channel and the likely cause (nothing was sent, a campaign paused, a page lost clicks), and put the action for that cause in Step 3.

If conversion moved

You are answering two questions: where in the journey did people stop, and which device did it?

1. Shopify: Conversion rate breakdown. Shopify’s help describes this as the funnel from all sessions to sessions with cart additions, sessions that reached checkout and sessions that completed checkout, and says each step’s rate is that step’s sessions divided by total sessions. Compare this week’s steps with last week’s. The step where the rate fell most is where to look.

These are hypotheses, not diagnoses: the data tells you the step, and a test on the live store tells you the cause.

2. Shopify: Sessions by device. Shopify’s help lists it as showing the types of devices used to access the store. Shopify’s help does not say whether conversion rate can be split by device, so work it out yourself: take orders by device from a custom Shopify exploration or from GA4 if you can, and divide by that device’s sessions. A drop on mobile alone points to something that behaves differently on a phone; mobile conversion rate on Shopify lists the common causes.

3. Shopify: Sessions by landing page. If the drop is concentrated on one entry page, the problem may be that page or the traffic sent to it.

4. Check the filter and the baseline. Confirm the Human or bot session filter matches the weeks you are comparing with, and that you are comparing only weeks after the September 2026 change. A “drop” that coincides with the rollout may be a measurement change, not a store problem.

5. Try the thing. Place a test order on a phone and walk the step that fell, on the device that fell. If you changed anything last week (a theme edit, a new app, a discount, a shipping rate), that is the first suspect, because it is the cheapest to check.

Decision for the conversion branch: name the step and the device, name the most recent change on the store, and make the action a check of that change.

If order value moved

1. Shopify: Orders over time. It shows average units per order and average order value alongside the order count (Shopify Help Center, order reports). Did units per order change, or price per unit?

2. Shopify: Sales by discount codes. Shopify’s help describes this as sales grouped by discount name or code, and says it helps you see how often your discount codes are applied. A new code, a code that leaked to a coupon site, or an automatic discount that now stacks can pull average order value down while orders rise.

3. Shopify: Items bought together. It shows the most common product combinations purchased together (processed orders only, per the help). If your best bundle stopped selling, or a usual pairing went out of stock, order value can fall without any price change.

4. Your own changes. Check whether the free shipping threshold, a bundle, an upsell app or a price changed, and check any such change against margin before you keep it. A rising average order value is not automatically good news: if it rose with a price increase and orders fell, net sales may have gone down.

Decision for the order-value branch: find the lever that changed (a code, a threshold, a bundle, stock) and decide whether to reverse it, keep it or test it.

Step 3 (minutes 25 to 30): one action, written down

The review earns its keep here. Choose one change and write four things in the log:

1. What you will do. Specific enough that someone else could do it. “Place test orders on a phone with each payment method, and roll back the app installed on Tuesday if any step fails” is an action. “Improve checkout” is not.
2. Who does it, and by when.
3. Which number should move, and in which direction.
4. When you will look. Put the date in your calendar. One week is often too short to judge a store change; look again in two to four weeks and compare against the range you wrote down, not your memory.

Three rules keep the action honest:

• One change at a time per branch. Change the offer, the checkout and a campaign in the same week and you will not know what worked.
• Cheap and reversible first. Checking, testing and rolling back come before redesigning.
• “Do nothing” is allowed, but write it down with the reason (“inside the usual range, nothing to explain”).

The worked example, finished

Back to the illustrative store. Conversion was the branch. Its device split (still invented) looks like this:

Almost all of the lost orders are on mobile: 12 fewer mobile orders against two fewer on desktop. Say the Conversion rate breakdown shows add-to-cart and reached-checkout rates roughly steady but fewer sessions completing checkout, and the one change on the store last week was an app installed on Tuesday. The step is checkout and the device is mobile.

The log entry reads: Place test orders on a phone with every payment method today. If a step fails, remove the Tuesday app and retest. Mobile checkout completion should recover towards the four-week range. Check at next Tuesday’s review. That is a cheap, reversible, single action tied to a number.

What stays out of the weekly review

These matter, but they move too slowly or are too noisy for a weekly check, and belong in a monthly review:

• Customer lifetime value, repeat purchase rate and cohorts. They need months of orders to mean anything.
• CAC payback and contribution margin. They need a full month of spend and costs.
• Individual keyword rankings. They move daily for reasons unrelated to your store. Weekly non-brand clicks in Search Console tell you more.
• Per-ad and per-creative detail. That belongs to whoever runs the ad account.

Key takeaways

• Review in a fixed order: orders and net sales, then the branch that moved, then one written action.
• Net sales are roughly orders times order value, and orders are sessions times conversion rate, so there are three branches: traffic, conversion and order value.
• Judge each week against the last four weeks’ range and the typical swing for your order count. As an illustration, a 40-order week swings by about 6 orders by chance.
• Shopify’s session measurement update (21 to 23 September 2026) changed how sessions and conversion rate are counted; orders and sales were not affected. As at 31 October 2026, treat conversion rate as a new baseline from that date.
• Open Shopify reports first. Use GA4, Search Console and Klaviyo when the traffic branch calls for them, and on Tuesday or later, because Google says GA4 processing can take 24 to 48 hours.
• Make the action specific, cheap and reversible, tied to a number, with a date to check it.

FAQs

What should I review in my Shopify store every week?

Review orders and net sales first, then whichever of sessions, conversion rate or average order value moved outside its usual range, then one change you will make. Add a standing check on quiet weeks. Customer lifetime value, cohorts and CAC payback belong in a monthly review.

Which Shopify report shows conversion rate?

Conversion rate over time shows the percentage of sessions that resulted in a purchase, grouped by day, week or month. Conversion rate breakdown shows the funnel from all sessions to cart additions, reached checkout and completed checkout. Both are in Analytics > Reports, according to Shopify’s help. Since 21 September 2026, compare only weeks after the session measurement update.

Why did my Shopify conversion rate change in late September 2026?

It may be a measurement change, not a store change. Shopify’s session measurement update (21 to 23 September 2026) changed how sessions are counted, and Shopify says conversion rate can shift even when orders and sales do not. Treat it as a new baseline, and check orders and net sales over the same dates to see whether the store itself changed.

Do I need GA4 and Search Console for a weekly review?

Not every week. Shopify’s reports cover orders, sessions, conversion rate and order value. Open GA4 Traffic acquisition and Search Console when the traffic branch moved and you need to know which channel changed, and for the rotating checks on quiet weeks.

How is this different from watching three numbers?

The three numbers post says which metrics deserve a permanent place on your wall. This routine says what to do with your screen open: what to check first, how to find the branch that moved, and when to stop. Use them together.

Keep reading

• The DTC growth system: paid, organic, CRO and email
• Shopify growth stalled: diagnose in order
• Why your Shopify conversion rate dropped

Not sure which branch to fix first?

Running the review every week but unsure which fix comes first? A weekly review is only as good as the numbers it reads, and checking those numbers is where our work starts. Book a free 30-minute diagnostic: we look at where your store loses buyers after the click, using your own numbers, and tell you what to fix first. We do not run ads or email. No pitch. No email gate.

If your weekly numbers keep pointing at conversion or tracking, that is the work in our Shopify CRO service.