Free Shopify store auditSpeed, SEO and conversion leaks — no cost, no obligation.
Claim it
Thriftizer Solutions LLPShopify Select Partner
Book a Growth Audit
Shopify Nov 21, 2026 9 min read

Shopify's Store Activity Log: Who Changed the Discount During Your Sale

Shopify rebuilt the Store activity log on 13 November. Here is where to find it, what it records, the discount and product changes it won't show you, and how to run a BFCM-to-Boxing-Day code freeze around the gap.

Shopify's Store Activity Log: Who Changed the Discount During Your Sale

Open a new tab, go to Settings → General in your Shopify admin and look for the Store activity log link near the bottom of the page. That is the shortest answer to the question most people arrive with. It is a running record of admin-level events: who signed in, who was invited, who changed a permission, which app was installed, when the theme was published, when someone touched your payout details. Shopify rebuilt it on 13 November, which lands about two weeks before the only stretch of the year where a Canadian store genuinely cannot afford a mystery change.

The less welcome part: the Shopify store activity log will not reliably tell you that a discount went from 20% off to 40% off at 9:40pm on Black Friday, or who did it. Knowing where the gap is matters more than knowing where the link is.

Who can see the log, and what to check before you need it

The store owner sees it. Staff with full admin permissions generally do. Limited staff do not, and that is the correct setup. If you are a collaborator working on a client store, assume you may not have access and ask the owner to pull entries rather than going hunting.

Do this now rather than during an incident: open the log, scroll to the oldest entry it will load, and note the date. Retention is not infinite and it is not something to discover at 2am on 26 December when you are trying to prove a change happened five weeks earlier. If the log only goes back a few weeks, your real audit trail for peak season has to be something you build yourself, before the season starts.

What the rebuilt store activity log records well

The things that are genuinely useful in it, in roughly the order they have saved us grief:

  • Staff logins, including failed ones and logins from new locations. Useful when an account has been shared and nobody will admit it.
  • Staff invitations, permission changes and removals. This is the one that matters most in December.
  • App installs and uninstalls, plus the permission scopes granted. An app installed at 11pm on Cyber Monday by someone trying to fix a problem is a very common root cause.
  • Theme publishing. Not the code edit, the publish. Different thing, and the distinction will come up later.
  • Plan changes, domain changes, and changes to banking and payout information.

The rebuild made it faster to filter by person and event type, which is the difference between useful and ornamental. Before, you scrolled. Now you can narrow to one staff member across a date range and actually read the result.

What it does not record, including the discount

Here is the honest shape of the gap. The activity log is an account and settings log, not a content log. It does not give you field-level before-and-after values on products, variants, prices, inventory, metafields, or discount codes. So the question "who changed the discount during our sale" is usually not answerable from it directly.

Theme code is a partial exception. The Shopify code editor keeps older versions of files edited through the admin, so if someone hand-edited a snippet during the freeze you can often see the prior version and roll back. If your theme is connected to GitHub, your commit history is the real record and it is better than anything native. If someone edited through the theme editor instead, you get the published-theme event in the activity log and nothing more granular.

For products and discounts, the Admin API exposes an events feed per resource that a developer can pull. It gives you timestamps and short descriptions of what happened, which is enough to narrow a window, though it is not a clean diff of old value versus new value. Treat it as forensic material, not as a report.

Reconstructing a discount change from your orders

When the log can't tell you, the orders can. Go to Orders, filter by discount code, set the date range, export to CSV. Every order carries the discount amount that was actually applied. The moment the applied rate jumps, you have your window, and the window is what you take back to the activity log to see who was signed in.

A worked version. Say a code was meant to run at 20% off and was sitting at 35% for three hours and ten minutes on the Saturday after Black Friday. Your export shows 214 orders in that window with an average pre-discount cart of CAD 128. Intended discount per order: 128 × 0.20 = 25.60. Actual: 128 × 0.35 = 44.80. Difference of 19.20 per order, times 214 orders, is 4,108.80 in margin you did not plan to give away. On a 42% gross margin that cart was contributing 53.76 per order; after the overage it contributes 34.56. You are still above water, barely, and now you know it rather than finding it in the January P&L.

Run that arithmetic before you start the conversation about who did it. The size of the number decides whether this is a quiet word or a process change.

Tracking who edited a product when Shopify won't

Three options, in increasing order of effort.

Shopify Flow. Free on Plus and available to most merchants now. Build a workflow on the product-update trigger that posts to Slack or appends a row to a Google Sheet with the product title, the new price and the timestamp. It will not always name the person, but it gives you a timeline you own, with retention you control, and it takes about twenty minutes to set up. Do the same for discount creation if your plan supports the trigger.

A backup and version-history app. The category exists precisely because the platform does not do this. These capture snapshots and diffs of products, collections, themes and sometimes discounts, and let you restore. If you have a catalogue in the thousands and more than three people with edit rights, this is cheap insurance.

A webhook listener. Subscribe to products/update and discounts/update, write the payload to your own store, done. We build these for merchants who need a real audit trail for finance or compliance reasons, and it is a small piece of custom Shopify app work rather than a project. The payload tells you what the record looks like after the change; pair it with the previous stored snapshot and you have your diff.

Set staff permissions before the sale, not during it

The fastest way to not need an audit trail is to reduce the number of people who can cause the problem. Two weeks before BFCM, we go through the staff list on every store we run and do this:

  • Remove anyone who has left. Agencies, freelancers, the summer co-op student, the dropshipping consultant from 2023. Collaborator accounts linger for years.
  • Strip Edit discounts and Create discounts from everyone who does not need it for their job that week. Customer support does not need it. Your paid-media contractor absolutely does not need it.
  • Same for Edit products if your catalogue is settled. A content person can draft in a staging theme or a spreadsheet for a fortnight.
  • Enforce two-step authentication on every account with product, discount or settings access. Shopify lets you require it at the organisation level.
  • Name one person who holds discount permissions and one backup. Two people. Write their names in the runbook.

When something does go wrong, a two-name list turns a forensic exercise into a phone call.

The code freeze: Black Friday to Boxing Day, not Black Friday to Cyber Monday

Canadian stores run a longer peak than most people plan for. Black Friday and Cyber Monday, then a steady December, then Boxing Day and the week after, which for a lot of apparel and home brands is the single biggest trading day of the year. Freezing from the Friday before BFCM week through 27 December is the version that actually matches the revenue curve.

What the freeze covers: no theme publishing, no app installs, no checkout or shipping profile changes, no tax settings, no payment method changes, no bulk product edits, no new discounts outside the planned calendar. What it does not cover: inventory updates, order fulfilment, customer support actions, and pausing a discount that is losing money.

The one exception worth writing into the policy ahead of time is carriers. After the Canada Post disruption of recent years, every store we work on keeps a second carrier configured and tested in shipping settings before the freeze starts, so that switching to it during the freeze is a pre-approved action rather than an emergency change. Test it in November. A rate change made in a panic on 18 December, with nobody recording why, is exactly the kind of thing you will be reading the activity log to understand in January.

The other Canadian landmine is tax. If someone toggles a tax override or a product's taxability mid-peak, the error shows up differently by province: 13% HST in Ontario, 5% GST in Alberta, GST plus PST in BC and Saskatchewan, and QST in Quebec. Under-collection is not recoverable from the customer after the fact, and a few hundred orders across provinces is a tedious remittance problem. Tax settings belong in the frozen category, no exceptions. The same goes for French content if you sell into Quebec, where bilingual product and checkout copy is a Bill 96 obligation rather than a nice-to-have. Someone "tidying up" translations during the freeze can create a compliance problem and a conversion problem in the same afternoon.

Agency and staff accountability in a shared admin

If an agency works on your store, the activity log is the thing that keeps everyone honest, in both directions. We ask clients to check it. Collaborator access is scoped, every change we make during a freeze window goes through a written approval in the shared channel first, and the log should match that record line for line. If it doesn't, somebody has gone around the process and it is better to find out in week one of December than in February.

Merchants running more than one store, or a Canadian storefront plus a US-facing one for cross-border orders, should be looking at this at the organisation level rather than store by store. Different stores usually have different staff lists, and the one you look at least is the one with the stale permissions. Our Canada team runs this review as part of pre-peak prep, and it typically turns up two or three accounts that should have been removed months ago.

Exporting the log, and why a screenshot beats an export

There is no one-click CSV export of the store activity log in the admin. If you need entries preserved, your options are to copy the rows you care about into a document at the time, or to have a developer pull the relevant events from the Admin API into your own storage on a schedule.

Practically: the moment you spot something odd, screenshot the log with the timestamp and the staff name visible, before anyone starts fixing things. Evidence first, remediation second. We have lost the trail more than once by rushing to repair the discount and only thinking about proof afterwards, by which point the entries had scrolled past the point where filtering found them quickly and the person involved had a different recollection of the timing.

Pair that screenshot with the orders CSV for the affected window. Those two artefacts together are enough for a chargeback dispute, an insurance claim, a supplier conversation, or just a calm Monday meeting.

What to do this week

Open Settings → General, find the store activity log, confirm you can read it and note how far back it goes. Then open your staff list and remove the people who should not be there. Those two tasks take under an hour and cover most of the risk.

If you want a second pair of eyes on permissions, app scopes and the state of your theme before the freeze, our free store audit covers it, and we will tell you plainly if the answer is that you are already fine.

Previous postNext post

Ready to scale your D2C brand profitably?

Let's build a growth engine that drives more traffic, more conversions and more profit.

Book a Growth Audit
📅 Free Audit💬 WhatsApp