Why Is My Google Ads
CPA So High?

Low Quality Score — you're paying a CPC penalty

Quality Score (1–10) directly multiplies your effective CPC. A QS of 4 means you pay up to 400% above the minimum bid to win auction positions. A QS of 9 means you pay 30–40% below market rate. Since CPA = CPC ÷ CVR, a QS penalty compounds directly into CPA.

How to diagnose: add the Quality Score column to your Google Ads keywords view. Any keyword below 5 is costing you significantly more than it should. Below 3 warrants immediate pausing or restructuring.

Root causes of low QS: ad copy that doesn't match keyword intent, landing page that doesn't match ad copy, slow landing page load time (>3 seconds), or low historical CTR on those keywords.

Landing page CVR below 2% — the most impactful fix

CVR is the denominator in the CPA equation: CPA = CPC ÷ CVR. A 1% CVR at $5 CPC = $500 CPA. The same CPC at 5% CVR = $100 CPA. No amount of bid optimisation, keyword pruning, or audience refinement will compensate for a fundamentally low-converting landing page.

Industry benchmarks for context: Google Search average CVR is 4.8% across all industries. Below 2% on a search-intent landing page almost always indicates a structural problem: form friction, page load speed, offer-message mismatch, or trust signals missing.

The highest-impact single change most advertisers can make: reduce form fields to the minimum required. Every additional field drops CVR by 10–15%. If you're asking for phone + email + company + name + message on a first-touch form, you're likely converting at 0.5–1%.

Broad keyword match — irrelevant traffic inflating CPA

Google's broad match algorithm has become more aggressive since 2022. A broad match keyword like "project management" now triggers for queries like "to-do list app", "team collaboration tools", "Asana vs Monday", and "what is agile methodology" — only some of which indicate buying intent. Each irrelevant click is spend without conversion potential, which raises average CPA across the campaign.

How to diagnose: pull your Search Terms report for the last 30 days. Sort by cost. Identify queries with high spend and zero conversions. These are the broad match leakage. Add negatives for irrelevant terms and convert high-spend terms to phrase or exact match.

Brand/non-brand mixing — averaging out the numbers

Brand keyword CPA is typically 3–5× lower than non-brand CPA. If you're running brand and non-brand keywords in the same campaign, your blended CPA looks reasonable but hides that non-brand performance may be significantly above your break-even.

Example: brand keywords convert at 15% (CPA = $33 at $5 CPC). Non-brand keywords convert at 3% (CPA = $167 at $5 CPC). Blended CPA if 50/50 traffic mix: $100. The $100 blended number masks that non-brand is unprofitable while brand is highly efficient.

Always separate brand and non-brand into distinct campaigns. This also prevents Smart Bidding from learning the wrong signal — a high brand CVR skews the algorithm's understanding of what good performance looks like for non-brand terms.

Smart Bidding launched too early — algorithm optimising on noise

Target CPA and Target ROAS bidding require a minimum of 30–50 conversions per month to function reliably. Below this threshold, the algorithm is making bid decisions based on statistically insignificant data — one or two conversion events can swing the model significantly, producing erratic CPCs and elevated CPAs.

The symptom: CPA is volatile week-to-week with no clear trend despite no changes in targeting or creative. The algorithm is confused because it doesn't have enough signal.

Solution: use Maximize Conversions (no target) for the first 30+ conversions to let the algorithm learn without a CPA constraint. Only introduce a Target CPA after you have a stable baseline — set it 20–30% above your observed CPA initially, then tighten gradually over 4–6 weeks.

Conversion tracking issues — phantom high CPA

If your conversion tracking is broken or double-counting, everything you think you know about CPA is wrong. This is more common than most advertisers assume — a site update, CMS change, or tag manager misconfiguration can silently break conversion recording without any visible error.

Symptoms of under-tracking: CPA suddenly spikes with no campaign changes; Google Ads conversion count diverges significantly from actual revenue in your CRM or Shopify. Symptoms of double-counting: CPA looks unusually low; conversions exceed actual orders.

Diagnostic: compare Google Ads reported conversions against your backend (Shopify orders, CRM closed deals, GA4 transactions) for the same period. If the numbers diverge by more than 15%, investigate your conversion setup before making any campaign changes.