Why Shopify Sales Drop 20% Overnight (And How to Actually Find Out Why)

If your Shopify sales fell sharply and you're staring at a dashboard trying to work out why, the honest first answer is: it's rarely one thing. A 20% drop usually comes from two or three smaller problems overlapping at once, and the instinct to "just run more ads" is almost always the wrong first move, because more traffic into a broken funnel just costs more money for the same result.

Here's the order to actually check things in.

1. Separate traffic from conversion

Before anything else, split your drop into two numbers: how many people visited, and what percentage of them bought.

• If sessions fell and conversion stayed flat, this is a traffic problem: a channel slowed down, an ad account got flagged, a seasonal dip, or an SEO ranking change.
• If sessions stayed flat but conversion fell, this is a funnel problem: something changed on your site, in your offer, or in who's arriving.
• If both fell, you likely have two separate issues at once, and fixing only one won't fully explain the drop.

Most sellers skip this step and go straight to guessing. It's the single most useful five minutes you can spend.

2. Check whether your traffic mix has shifted

Even with stable total sessions, the source of that traffic matters enormously. Paid social typically converts lower than email, direct or organic search. If a channel that converts poorly grew as a share of your traffic (a new ad campaign, a viral post, a bot spike), your blended conversion rate can fall even though nothing on your site changed.

Check conversion rate by source, not just overall. A rough rule: if conversion fell across every source by roughly the same amount, look at your site. If it only fell overall because of a mix shift, look at your traffic sources.

3. Look at mobile specifically

Most ecommerce traffic is mobile, often 65 to 75%. A change that only affects mobile, a slower-loading app you installed, a review widget, a new tracking pixel, can tank your overall numbers while desktop looks fine. Compare mobile and desktop conversion separately before concluding the whole site has a problem.

4. Check anything you changed in the last 4 to 6 weeks

Sales drops are rarely caused by nothing. Go through what actually changed:

• Shipping thresholds or shipping costs
• Any new app or widget added to product or checkout pages
• Pricing or discount code changes
• Checkout steps added or removed
• A theme update

Cross-reference the date of the drop against the date of each change. This is the single fastest way to find a likely cause, because coincidence in timing is a strong (though not certain) signal.

5. Rule out the boring explanations

Before assuming something is broken, check:

• Stock: is a best-seller out of stock or has its variant availability changed?
• Seasonality: compare to the same period last year, not just the last 30 days.
• Payment processing: a failed payment gateway update can silently block a percentage of checkouts.
• Tracking: a broken pixel or analytics tag can make sales look lower than they are, when really you're just measuring wrong.

6. Don't fix five things at once

Once you have a shortlist of likely causes, resist changing everything simultaneously. If you fix the shipping threshold, speed up your product page, and change your ad budget in the same week, you'll have more sales (hopefully) but no idea which fix actually mattered, which means you can't repeat the win or avoid the mistake next time.

Sequence it: fix the most likely cause first, measure for a week, then move to the next.

A structured way to think it through

If you'd rather not do this analysis alone, this is exactly the kind of problem that benefits from being looked at from a few different angles before you act: is it really the funnel, or is it traffic quality? Is the shipping change actually the cause, or a coincidence? Should you hold ad spend or test a small amount anyway?

MultiDI is built for exactly this kind of decision. You describe what's happened, and it puts the question through multiple independent AI perspectives that research the likely causes, challenge each other's assumptions and disagree where the evidence is genuinely unclear, before giving you a conclusion and a practical next step. It won't replace looking at your own data, but it's a useful way to pressure-test your own theory before you spend a limited marketing budget acting on it.