Revenue is down. Google Ads is spending. Meta is spending. So the ads must be the problem.
Sometimes they are.
But an eCommerce sales drop can start almost anywhere: fewer people may be looking for the product, an important variant may be out of stock, checkout may be failing on mobile, shipping may have become more expensive, organic traffic may have fallen, or analytics may simply be reporting fewer purchases than the store actually processed.
If you start changing campaigns before you know which part of the business changed, you can create a second problem while leaving the first one untouched.
Our preferred approach is simpler: start with the revenue decline itself and work backwards.
First confirm that sales actually fell. Then identify whether the change came from traffic, conversion rate, average order value, or post-purchase adjustments. After that, narrow the problem by channel, device, market, product, and funnel stage.
Only then does it make sense to decide whether advertising caused the decline.
Start by locating the sales drop
For a first-pass diagnosis, eCommerce revenue can be broken into three variables:
Revenue = sessions × conversion rate × average order value
If revenue falls, at least one part of that equation has changed.
For example:
- If sessions fall while conversion rate and average order value remain similar, investigate acquisition.
- If traffic stays stable but fewer visitors place an order, investigate conversion.
- If order volume stays similar but revenue falls, investigate average order value and product mix.
That equation describes gross sales well enough for diagnosis, but it is not the whole financial picture. Refunds, cancellations, failed payments, returns, discounts, and taxes can all affect the final numbers you see in different reports.
This is why the first question should not be “Why are sales down?”
It should be:
“Which number changed?”
1. Confirm that the drop is real
Before analysing campaigns, compare advertising dashboards with the system that actually records orders.
A tracking problem can look almost identical to a sales problem.
A purchase event may stop firing after a checkout update. Revenue values may be missing. Transactions may be duplicated. Consent changes can affect measurement. One reporting platform may also attribute a purchase differently from another.
GA4 does not automatically collect a complete eCommerce funnel in a standard implementation. Events such as add_to_cart, begin_checkout, and purchase need to be sent with the appropriate data. Some commerce platforms can handle part of this integration automatically, but the implementation still needs to be checked.
Start with your commerce platform or order management system and compare:
- Completed orders
- Gross and net sales
- Average order value
- Discounts
- Refunds and cancellations
- Failed or unpaid orders
- New and returning customer orders
Then compare those figures with GA4, Google Ads, Meta Ads, and any internal reporting dashboard.
Do not expect every platform to show exactly the same revenue.
Different systems use different attribution models, reporting windows, time zones, customer identification methods, and definitions. What matters at this stage is whether the store actually processed fewer valuable orders.
Also check the comparison period before drawing conclusions.
An incomplete Tuesday should not be compared with a completed Monday. A normal week should not automatically be compared with Black Friday. For a seasonal business, the same period last year can provide useful context, but promotions, stock, pricing, and product mix still need to be comparable.
2. Find the part of the revenue equation that changed
Once you know the revenue decline is real, separate traffic, conversion rate, and average order value.
This immediately reduces the size of the investigation.
If traffic fell
Do not look only at total sessions.
Break the change down by acquisition source:
- Paid search
- Paid social
- Organic search
- Direct
- Referral and affiliate traffic
Then segment further by device, country, landing page, campaign, and product category.
A total traffic chart can hide a very specific problem. Desktop may be stable while mobile traffic is down. Google Ads may be unchanged while organic traffic has fallen sharply. One category that used to generate a large share of sessions may have disappeared from search results.
Finding the affected segment tells you where to look next.
If conversion rate fell
Stable traffic with fewer orders points somewhere inside the shopping journey.
Break the funnel into stages rather than looking only at the overall conversion rate:
- Product view → add to cart
- Add to cart → checkout
- Checkout started → payment
- Payment → completed order
Then compare those stages by device, browser, country, traffic source, and major product category.
Imagine overall conversion falls from one week to the next, but add-to-cart behaviour has barely changed. The sharpest decline happens between checkout and completed purchase, and it is concentrated on mobile.
That evidence points you toward checkout, payment, shipping, or a mobile-specific technical problem. Rebuilding product pages or changing ad creative would be a distraction.
If average order value fell
Revenue can decline even when order volume does not.
Check whether customers are buying:
- Cheaper products
- Fewer items per order
- More heavily discounted products
- Different product categories
Also check whether a bundle ended, free-shipping rules changed, cross-sells disappeared, or expensive bestsellers went out of stock.
Paid campaigns can still report a similar number of purchases while sending more traffic toward lower-value products. Looking only at cost per purchase would miss that change.
3. Check whether customer demand changed
Sometimes nothing is technically wrong with the website or campaigns. Fewer people simply want the product right now.
Demand can change because of seasonality, economic conditions, new competing products, trends, weather, events, or changing interest in the category.
Search Console and Google Trends are useful here.
For organic search, compare impressions and clicks for the queries and pages that normally generate meaningful traffic. Then look at countries and devices to see whether the decline is broad or concentrated.
Google Trends can provide another piece of context. If search interest for an important category has fallen across the market, your traffic decline may not be unique to your store.
Look for the same pattern across several channels:
- Did organic search impressions decline?
- Did paid search impression volume also fall?
- Are fewer people visiting directly?
- Is the decline concentrated in one category or country?
- Did something similar happen at the same time last year?
One signal is not enough to prove a demand decline. Several independent signals moving in the same direction make that explanation more credible.
Lower demand changes the decision you need to make.
The answer may involve forecasting, merchandising, budgets, pricing, product positioning, or creating demand for another category. Rewriting ads alone cannot create search volume that no longer exists.
4. Check what customers can actually buy
Advertising performance depends heavily on what happens after the click.
Start with the products that generate the most sales or receive the most advertising spend.
For each one, check:
- Stock availability
- Important sizes, colours, or variants
- Current price
- Discounts
- Shipping availability
- Delivery time
- Final checkout cost
A product can still be listed as “in stock” while its most popular size is unavailable. A campaign can continue sending traffic to a bestseller that customers can no longer purchase in the configuration they want.
For Google Shopping and Performance Max, also inspect Merchant Center.
Google compares product data with information on the landing page. Price or availability mismatches can result in products being disapproved, which means they stop appearing in eligible Google placements until the problem is resolved.
Do not assume the product feed is healthy simply because the campaign itself is active.
Check whether the offer changed
A conversion decline can also come from a commercial change rather than a technical one.
Look at what was different immediately before the decline.
For example:
- A promotion ended
- Prices increased
- A free-shipping threshold increased
- Shipping became more expensive
- Delivery times became longer
- A popular bundle was removed
- A high-performing bestseller went out of stock
Also compare competitors where it is commercially relevant. A campaign can become less profitable without anything changing inside the ad account if the offer behind the ad becomes less attractive.
In that case, a rising cost per purchase may be evidence of the problem rather than its cause.
5. Follow the customer through the website
If traffic is stable and the commercial offer has not materially changed, look closely at the website.
Start by checking what changed around the same time as the sales decline:
- Website releases
- Theme or component updates
- New apps or third-party scripts
- Checkout configuration
- Payment providers
- Shipping rules
- Consent and tracking changes
Then test the buying journey rather than simply opening the homepage and declaring the site “working.”
Use the devices, locations, and products affected by the decline.
- Open an important campaign or SEO landing page.
- Browse to a product.
- Select a common variant.
- Add it to the cart.
- Apply a current promotion if relevant.
- Enter the target shipping location.
- Continue through checkout.
- Test the main payment method.
- Complete a test order where possible.
- Confirm that analytics and advertising events were recorded correctly.
A technical problem can affect only one browser, one payment method, one country, or one device. Looking only at site-wide averages can make it easy to miss.
Follow the funnel evidence.
If product views and add-to-cart rates remain normal but checkout completion suddenly falls, investigate checkout before redesigning the product page.
Check performance, but do not treat it as the answer to everything
Page performance is another useful signal.
Core Web Vitals can show changes in loading speed, interaction responsiveness, and visual stability. A release or new third-party script that makes important templates significantly slower deserves investigation.
But poor Core Web Vitals do not automatically explain a revenue decline. They are one part of the evidence.
The question is whether the performance change appeared on the same pages, devices, and period where conversion also deteriorated.
6. Compare the other acquisition channels
Paid advertising is only one source of customers.
Total revenue can fall while Google Ads and Meta Ads remain relatively stable because another large acquisition or retention channel weakened.
For organic search, compare:
- Clicks
- Impressions
- Important queries
- Landing pages
- Countries
- Devices
A broad organic decline may be related to technical problems, indexing changes, ranking changes, seasonality, or changes in search demand. Search Console is more useful when you segment the drop than when you look only at the total graph.
For email, compare the steps separately:
Messages sent → delivered → clicked → sessions → purchases
If fewer emails were sent, email revenue may fall even when each campaign performs normally.
If clicks remain stable but purchases decline, the problem is probably further down the journey.
The same logic applies to affiliates, referral traffic, direct traffic, and returning customers.
The goal is not to prove that advertising is innocent. It is to understand how much of the total sales decline actually came from paid media.
7. Now inspect Google Ads and Meta Ads
Once you know what happened elsewhere in the business, the advertising data becomes much easier to interpret.
Compare the period before and after the decline and look at the chain from delivery to revenue.
Depending on the campaign type, useful metrics include:
- Spend
- Impressions and reach
- Clicks
- Cost per click
- Click-through rate
- Purchases
- Conversion rate
- Cost per purchase
- Conversion value
- Revenue or ROAS
The useful question is not simply whether ROAS went down. Ask where the change started.
If impressions or traffic fell
Check campaign status, budgets, targeting, product eligibility, bids, search demand, policy issues, and any recent account changes.
Google Ads itself lists recent campaign changes, budget and bidding settings, conversion tracking, targeting, policy status, and billing among the areas to investigate when performance changes.
If traffic remained stable but purchases fell
Compare the campaign conversion rate with the store's overall conversion rate.
If both declined at roughly the same time, the cause may sit on the site or in the offer rather than inside the campaign.
If only paid traffic deteriorated, inspect landing-page alignment, traffic quality, audience changes, search terms, placements, creative, and campaign structure more closely.
If purchase volume stayed similar but revenue fell
Look at which products and customers the campaigns are bringing in.
The account may have shifted toward cheaper products, heavier discounts, different markets, or lower-value customers.
Check what changed before the decline
Review account history and product data for meaningful changes such as:
- Budget changes
- Bid or optimisation changes
- Campaign restructuring
- Targeting changes
- New or removed creative
- Conversion tracking changes
- Landing-page changes
- Product feed changes
- Product disapprovals
Timing does not prove causation, but it gives you a useful hypothesis to test.
Avoid changing several major variables at the same time simply because performance is down.
If budget, targeting, creative, campaign structure, and landing pages all change together, it becomes much harder to learn what actually improved or damaged performance.
8. Build a diagnosis before choosing the fix
By this point, you should be able to describe the decline more precisely than “sales are down.”
For example:
Revenue fell mainly because mobile conversion dropped. Traffic and average order value remained stable. The largest funnel change occurred between checkout and completed order, beginning immediately after a checkout release. Paid and organic traffic showed similar conversion declines.
That is a diagnosis you can act on.
Compare it with:
Meta ROAS is down, so we should launch new creatives.
The second statement describes a symptom in one reporting platform. It does not explain the business decline.
A useful diagnosis should answer:
- Is the revenue decline real?
- Did traffic, conversion rate, or average order value change?
- Which channels were affected?
- Which devices, countries, categories, or products were affected?
- Where did the biggest funnel change occur?
- What changed shortly before it?
- Do independent data sources show the same pattern?
You may still have more than one possible cause. That is normal.
The important difference is that you now have specific hypotheses that can be tested rather than a list of departments to blame.
A few patterns we look for during an eCommerce audit
The exact diagnosis depends on the store, but certain patterns quickly change where we focus our attention.
| What changed | What we investigate next |
|---|---|
| Store revenue is stable, but GA4 revenue dropped | Tracking implementation, purchase events, consent, transaction values, and reporting configuration |
| Traffic fell, conversion stayed stable | Acquisition channels, demand, SEO visibility, email volume, paid delivery, and referral sources |
| Traffic stayed stable, conversion fell | Funnel stages, products, pricing, website changes, checkout, payment, shipping, and traffic mix |
| Orders stayed stable, revenue fell | Average order value, product mix, discounts, bundles, stock, and higher-value product availability |
| Only mobile conversion fell | Mobile UX, performance, browser-specific problems, checkout, payment methods, and recent front-end releases |
| Paid and organic traffic both declined for one category | Demand, seasonality, stock, category changes, search visibility, and product availability |
| Ad clicks stayed stable, but paid conversion fell with site-wide conversion | The website, offer, checkout, product availability, pricing, and demand before making major campaign changes |
| Paid traffic declined while other channels stayed normal | Campaign delivery, budgets, bids, targeting, eligibility, account changes, feeds, creative, and platform-specific issues |
Fix the cause, not the dashboard that looks worst
Advertising dashboards attract attention because spend, revenue, ROAS, and cost per purchase are visible every day.
But ads do not operate independently.
They send people toward products that need to be available, correctly priced, competitive, and easy to purchase. They depend on accurate tracking. They depend on demand. And they depend on a website and checkout that can turn the visit into an order.
When sales fall, we recommend investigating in this order:
- Confirm the decline using order and revenue data.
- Separate traffic, conversion rate, and average order value.
- Identify the affected channels, products, devices, and markets.
- Check demand and seasonality.
- Review stock, pricing, promotions, shipping, and product data.
- Follow the customer through the website and checkout.
- Compare organic, email, retention, and other traffic sources.
- Analyse advertising within that context.
This process does not guarantee an instant answer. Real eCommerce systems have several variables changing at the same time.
What it does is reduce guesswork and prevent a healthy part of the business from being changed simply because it has the most visible dashboard.
Creative Brackets helps eCommerce teams investigate performance across marketing, analytics, UX, SEO, and development rather than looking at each area in isolation.
If sales have dropped and the reason is not obvious, our eCommerce audit can help identify where the change started, which findings matter, and what is worth fixing first.
