Why Your Ecommerce Ads Aren't Converting: Diagnostic Guide

You're getting clicks. Traffic is coming in. CPCs look reasonable. But nobody's buying. Or, worse — your ROAS was decent for months, and now it's fallen off a cliff and nothing you do in the ad account seems to fix it.

This pattern is everywhere in ecommerce. The reflex is to blame the ads: swap the creative, change the targeting, test new headlines, restructure the campaigns, switch agencies. Sometimes that works. More often, it doesn't — because the problem isn't in the ad account. It's somewhere between the click and the purchase.

Here's a diagnostic framework. Work through these in order. Each one is a potential explanation, and they compound — a store with two or three of these problems will see dramatically worse performance than the ad metrics would predict.

1. Your conversion tracking is broken

Start here, because every other diagnosis depends on accurate data. If your conversion tracking is miscounting, your ROAS numbers are wrong, and you're making decisions based on fiction.

The most common tracking problems:

Duplicate conversions. If you have both the Google Ads conversion tag and a GA4 import firing for the same event, you're counting every purchase twice. Check your Google Ads conversion settings — are there multiple conversion actions with "Include in Conversions" turned on? If so, you're inflating your numbers and the algorithm is optimizing for phantom conversions.

Missing server-side tracking. Browser-based tracking (the standard Pixel or gtag.js) misses 10–30% of conversions due to ad blockers, browser privacy settings, and iOS ATT. If you're running Meta Ads, Conversions API (CAPI) is a requirement, not an option. For Google, enhanced conversions or server-side GTM should be set up. Without these, your platform is under-reporting conversions, which means the algorithm has incomplete data and makes worse optimization decisions.

Cross-domain tracking gaps. If your checkout is on a different domain than your storefront (common with Shopify Plus custom checkouts, headless setups, or third-party payment processors), the user session can break. The click gets counted, the purchase doesn't. Check your GA4 Realtime report while making a test purchase to verify the session stays connected through checkout.

Attribution window mismatch. Google defaults to 30-day click attribution. Meta defaults to 7-day click, 1-day view. If you changed these settings at some point — or if your agency did — your ROAS data may not be comparable to historical performance. Check what windows are set and keep them consistent.

Run a test purchase through each channel at least once a month. Click your own Google Shopping ad, complete a purchase, and verify the conversion appears in the Google Ads dashboard within 24 hours. Do the same for Meta. This takes ten minutes and catches tracking issues before they corrupt weeks of data.

2. The post-click experience is the bottleneck

The ad's job is to get the click. The landing page's job is to get the sale. When clicks are cheap but conversions are absent, the landing page is usually the problem.

Page speed. A product page that takes 4+ seconds to load on mobile loses half its visitors before they see anything. Run your product pages through PageSpeed Insights. Aim for a mobile score above 60. Common culprits: uncompressed images, too many Shopify apps loading scripts, heavy video embeds, custom fonts loading synchronously.

Mobile checkout friction. Over 70% of ecommerce traffic is mobile. If your checkout requires creating an account, entering a long-form address, or navigating multiple pages on a small screen, you're losing people. Shopify's Shop Pay accelerated checkout converts at nearly twice the rate of standard checkout for enrolled users. If you're not using it, you're leaving conversions on the table.

Product page trust signals. A new DTC brand with no reviews, no recognizable trust badges, and no social proof is asking a stranger to enter their credit card based on three product photos and a paragraph of description. Reviews, star ratings, user-generated content, and press mentions directly affect whether someone completes a purchase. If you have fewer than 10 reviews on your top products, that's a conversion problem — not an ad problem.

Price/shipping surprise. The second most common reason for cart abandonment (after "just browsing") is unexpected costs at checkout. If your product is $39 but shipping is $12, and the customer doesn't see that until checkout, they leave. Show shipping costs on the product page or offer free shipping above a threshold.

Ad-to-page mismatch. If your ad shows a specific product variant (black, size M) but the landing page opens on a different variant, the customer has to do work to find what they clicked on. This friction converts into bounce. Your ad creative and your landing page should show the same product, same color, same price, same offer.

3. You're targeting the wrong audience

"Wrong audience" doesn't always mean wrong demographics or interests. It often means wrong stage of the purchase journey.

Broad targeting with conversion objectives. If you're running Meta Advantage+ or broad interest targeting with a Purchase objective and a small daily budget ($20–50/day), the algorithm doesn't have enough conversion data to learn. It needs roughly 50 conversions per week to optimize properly. Below that, it's guessing. Either increase budget or switch to a higher-volume objective (Add to Cart) while you build data.

Google Shopping showing on irrelevant queries. Check your search terms report. If you're showing for queries that have nothing to do with your products — or for queries that are researchy rather than purchase-intent — your feed isn't optimized. Add negative keywords to your Shopping campaigns (yes, you can do this in Standard Shopping) and tighten your product titles to match the searches you want to capture. See our Google Ads ecommerce guide for the feed optimization framework.

Retargeting audiences are too small. If your retargeting pool is under 1,000 users, Meta can't deliver ads to it effectively. You'll see low delivery, high frequency, and declining ROAS. The fix isn't better retargeting — it's more prospecting to fill the top of the funnel.

4. Your ROAS was never real

Most operators don't want to hear this one. Sometimes ads appear to stop converting when they were never truly converting in the first place. The "good" ROAS you saw before was an attribution artifact.

Common scenarios:

PMax was capturing branded traffic. Your PMax campaign reported 8x ROAS. You increased budget. Then ROAS dropped to 3x. What happened? The initial high ROAS came from branded search queries — people who already knew your brand and would have bought anyway. When you scaled budget, PMax started prospecting with the additional spend, and prospecting ROAS is always lower than branded ROAS. The 8x was never real incremental performance.

Meta was running on 1-day view attribution. Meta's default attribution counts a conversion if someone viewed (not clicked) your ad and then purchased within one day. A chunk of those people would have purchased regardless — they saw your ad in their feed, scrolled past it, and bought later because they were already a customer or already in the purchase journey. When Meta changes its algorithm or your audience saturates, these "conversions" vanish because they were never caused by the ads.

Seasonal overlap. Your ROAS looked great during BFCM or a product launch. You assumed the campaigns were performing. Then it dropped back to baseline. That's not underperformance — that's the natural demand spike ending. If you set your "normal" ROAS expectations based on a peak period, everything afterward looks like failure.

Track blended ROAS alongside platform ROAS. If platform ROAS says 6x but blended ROAS says 2.5x, the gap is attribution inflation. See our ROAS benchmarks article for how to calculate and interpret blended ROAS correctly.

5. Nobody knows who you are

Every tactical problem above gets worse when your brand has zero awareness.

If your brand has zero awareness — meaning nobody in your target audience has ever seen your name, logo, or product before the ad — then every single impression starts from scratch. The ad needs to introduce your brand, build trust, explain the product, overcome objections, and close the sale. That's five jobs for one scroll-past moment in someone's feed.

The math on cold traffic to an unknown brand: 1–2% conversion rate on a good day. That means 98% of your paid clicks generate zero revenue. As CPMs rise (which they do every year), your CPA on cold traffic climbs and your ROAS falls. You can optimize creative, test landing pages, and restructure campaigns — and you'll get marginal improvements. But the structural ceiling is set by the fact that you're a stranger asking for someone's credit card.

The brands that maintain strong ROAS at scale are brands people have heard of. Their cold traffic isn't really cold. Their Shopping ads get clicked more because the logo is familiar. Their Meta ads convert better because there's pre-existing trust.

We cover this in our ecommerce brand building guide and in our ecommerce marketing strategy framework. The short version: if you've been running performance-only for 2+ years and your CAC keeps rising despite optimization, the problem isn't the ads. It's the brand.

Quick reference

Verify tracking — ten minutes of test purchases catches most data problems. Then audit post-click: page speed, mobile checkout, trust signals. Then check targeting: search terms for Google, audience overlap for Meta, retargeting pool sizes. Then question the baseline — was the "good" ROAS real, or was it branded traffic and seasonal overlap? Finally, check branded search volume in Google Search Console. If it's flat while ad spend has increased, you're buying cold traffic at scale without building any recognition.

If you've worked through all five and want help with the specifics, send us a brief. We'll tell you whether it's a tactical fix or a structural one — and whether we're the right fit to solve it.