Problem
Ecommerce conversion rate dropped suddenly: how to fix it
A conversion rate that falls sharply on a specific day almost always has one cause with a date attached. A rate that drifts down over weeks usually has several small ones. Work out which shape you have first, because a sudden drop is an investigation and a gradual one is a programme of work. For a sudden drop, the order of checks matters: measurement, then traffic, then the site itself.
By CartKernel · Last reviewed 2026-09-07
Does this look familiar?
- Conversion rate fell noticeably on one day and stayed down
- Sessions are flat or up while orders are down
- The fall is confined to one device, one browser or one country
- Order value held but order count dropped
- The drop coincides with a site release, an app install or a campaign launch
- Support has not received complaints, so the failure is silent
Causes, ranked
Why it happens, most common first
Check them in this order. The first two account for most cases we open.
- most common
The measurement broke rather than the store
A purchase event that stopped firing, a consent banner change, a duplicated session tag or a filter added to reporting can all halve a reported rate while orders continue normally. Checking store orders against analytics for the same days takes minutes and rules this out.
- most common
The traffic mix changed
A new prospecting campaign, a viral post, a broad match expansion or a bot wave brings sessions that were never going to buy. The denominator grows, the rate falls, and nothing about the store is worse. Segmenting by source usually shows the entire drop in one channel.
- common
Something in the buying path is broken
A payment method failing, a shipping rate returning nothing for certain addresses, a script error on the product page, or a required field that cannot be completed on one browser will stop a share of orders. These are silent, because people leave rather than report them.
- common
The best selling products went out of stock
When the products that carry most orders become unavailable, traffic keeps arriving from ads, email and search and finds nothing to buy. The rate falls immediately and recovers when stock returns, which makes it easy to identify in hindsight.
- common
A price, promotion or shipping change landed
A promotion ending, a delivery threshold being raised, a price increase or duties being added at checkout all change the offer. The effect is immediate and dated, and it is often made by someone who did not expect it to show up in a conversion report.
- occasional
Site performance regressed
A new app, an unoptimised image set or a third party script added to every page can slow the site enough to cost orders. This usually shows as a gradual decline, but a single heavy addition can produce a sharp one.
- occasional
Seasonality or an external event
A public holiday, a weather event, a competitor promotion or a category wide demand shift moves conversion without anything in your control changing. Comparing against the same period last year separates this from a defect.
The fix
In this order
Each step is something you can do today. Do them in sequence; skipping ahead is how a review fails twice.
Prevent it next time
- Alert on hourly orders falling below a threshold so silent failures surface within a day
- Run an automated checkout test against the live store on a schedule
- Keep one change log covering deployments, apps, pricing and campaigns with dates
- Check stock on the top revenue products before every campaign launch
Reconcile orders against analytics for the affected days
Count orders in the store admin and compare with recorded purchases in analytics for the same dates. If the store is normal and analytics is down, this is a tracking incident and the fix is in the tag, not the site.
Segment by source, device, browser and country
Break the rate down every way you can and find where the drop lives. A drop confined to one segment names its own cause: one channel points at traffic mix, one browser points at a script error, one country points at shipping or payment.
Read the change log for the exact date
List every deployment, theme edit, app install, price change, promotion end and campaign launch on and around the day the rate fell. Most sudden drops are explained by an entry in that list rather than by anything subtle.
Place a real order on the affected segment
Buy something on the browser, device and country where the drop is concentrated, using the payment method a customer would. Watch for errors in the console and note anything slow or confusing. This finds silent failures faster than any report.
Check stock on the products that carry the revenue
Identify the products that produced most orders before the drop and confirm they are available, priced correctly and reachable. Where they are out of stock, check what campaigns are still sending traffic to them.
Verify payment and shipping end to end
Compare payment attempts with successful charges in the provider dashboard, read the decline reasons, and test a checkout for each region and shipping method. A rate limit or a rule change on the provider side is a common silent cause.
Roll back the suspect change if you can
Where a release or an app is the likely cause, reverse it and watch the rate for a day rather than debating it. A quick reversal costs little and confirms or eliminates the theory faster than any analysis.
Write down what happened and add a monitor
Record the cause, the date and the fix, then add an alert that would have caught it sooner, such as an hourly order count threshold or a synthetic checkout test. The same class of failure tends to recur.
When to get help
Ask for help when the drop has lasted more than a few days, measurement is confirmed intact, and no segment isolates the cause, because at that point you need a systematic funnel analysis with session evidence rather than more hypotheses. It is also worth help when a payment provider or a third party service is implicated, since resolving those often requires reading provider logs alongside store logs and reproducing a fault that only appears for a subset of customers.
Free. We reply within one business day.
How large a drop is worth investigating?
Judge it against normal daily variation for your order volume. Stores with few daily orders see wide swings that mean nothing, while a high volume store should treat even a modest sustained change seriously. A useful rule is to investigate anything that persists for three days outside the range you normally see.
Could bot traffic be lowering my conversion rate?
Yes. Scrapers, monitoring services and automated traffic inflate sessions without ever buying. Check for sudden session growth from one source, one country or one browser signature with almost no engagement, and apply filtering so the reported rate reflects real visitors.
Should I compare against last week or last year?
Use both. Last week catches breakage and recent changes. Last year catches seasonality that would otherwise look like a decline. When they disagree, the year over year comparison is usually the more honest read for a store with a strong seasonal pattern.
My conversion rate dropped but revenue is flat. What does that mean?
Usually that sessions grew from a source that does not convert, while the buying customers behaved as before. Check whether a campaign, a piece of coverage or a bot wave added traffic. In that situation the rate is a mix effect rather than a problem to fix.
Services that fix it for good
Related problems, answers and terms
Stop the leak.
A free Growth Analysis finds what is broken and ranks it by the revenue it costs you.