Hey Sage · Free Meta ads diagnostic · AUD
Find the
actual leak
Meta Ads Performance Doctor for ecommerce brands
High CPA is a symptom, not a diagnosis. Enter the numbers from Ads Manager and your store to see whether the first place to look is delivery, creative, click quality, the landing page, checkout or unit economics.
Read the funnel in order.
A weak result at the bottom of the funnel can begin much earlier. High CPA might come from expensive delivery, an ad that does not earn a click, slow page loading, an offer that does not earn an add to cart, checkout friction, or a target CPA that the product economics simply cannot support.
The doctor models one connected path from impression to purchase. It deliberately avoids using platform ROAS alone because the platform can over-credit revenue and hide the stage where demand is leaking.
CPM and frequency tell you what it costs to reach the market and whether the same people are carrying too much of that delivery.
Link CTR tests whether the ad earns an intentional site visit. All-click CTR is not an equivalent input.
Landing-page views versus outbound clicks surfaces slow loads, accidental clicks and measurement gaps.
Add-to-cart rate is the first meaningful sign that the product, price, message and page are working together.
Initiated-checkout to purchase loss points toward shipping surprises, payment issues, trust or checkout usability.
The modelled CPA is held against your own target, not somebody else’s public benchmark.
Method note: the funnel maths comes entirely from the numbers you enter; the diagnostic order is Hey Sage planning logic, not a Meta health score. Reviewed July 2026 against Meta’s Performance 5 guidance and Conversions API overview. If measurement is the likely leak, use the Attribution Gap Diagnostic. If creative is the likely leak, build the next test batch in the Creative Testing Pipeline Planner.
What each diagnosis means.
Delivery pressure
High CPM by itself does not prove targeting is wrong. Premium markets, peak periods and valuable audiences can all cost more. Look for the combination of expensive reach, rising frequency and weakening click response before blaming the auction.
Creative and message
If link CTR is weak, the ad is failing before the website gets a chance. Test meaningfully different concepts—problem, proof, demonstration, comparison, founder, review—not twenty cosmetic edits of the same idea.
Click-to-page loss
A large gap between outbound clicks and landing-page views can come from page speed, accidental placement clicks, redirects, consent behaviour or event configuration. Check the actual mobile path before changing the campaign.
Landing page and offer
Healthy click response with weak purchase behaviour usually calls for work on message match, product education, proof, price framing, shipping, returns and the offer. It is rarely fixed by another interest audience.
Tracking confidence
If Pixel and Conversions API events are duplicated, missing values, using different currencies or materially disagreeing with Shopify, the bidding system may be learning from the wrong outcome. Fix measurement before optimising to it.
Meta ads diagnosis FAQ
Why do my Meta ads have clicks but no sales?
The click may be low intent, the page may load poorly, the ad and landing page may tell different stories, the offer may not justify the price, or checkout may introduce friction. Compare link CTR, landing-page-view rate, add-to-cart rate and purchase CVR in that order.
Is a high Facebook CPM always a problem?
No. CPM changes by audience, geography, season, objective and auction pressure. It becomes more concerning when it rises alongside frequency while link CTR and downstream conversion weaken.
What is the difference between link CTR and all-click CTR?
Link CTR counts clicks intended to leave Meta for your destination. All-click CTR also includes reactions, profile clicks and other interactions, so it can make an ad look more effective at driving traffic than it is.
What landing-page-view rate should I use?
Use your actual landing-page views divided by outbound link clicks for the same date range. A material gap deserves investigation, but there is no universal pass mark because browser behaviour, consent and event configuration differ.
Should I optimise campaigns before fixing tracking?
Usually not. If purchase events, value or currency are unreliable, both your diagnosis and Meta’s optimisation can be unreliable. Reconcile events with Shopify and check Pixel/CAPI deduplication first.
Can this tool tell me which ad to turn off?
No. It diagnoses the likely stage of the funnel. Ad-level decisions still require enough spend, conversion lag, concept-level analysis and comparison against the account’s own economics.
Want the account opened up properly?
Hey Sage can screen-share the structure, measurement, creative and store funnel—then show you what should change first and what should be left alone.
RUN BETTER ADS
