How to Read Your Waste Audit Report

Total cost across all search term activity in the last 30 days. This is your denominator — every percentage metric is calculated against this figure.

Pre-calculated by the waste management pipeline — not derived from a simple zero-conversion filter. Each dollar of waste is classified by entity type, waste category, and recommended action before the audit runs.

Wasted Spend expressed as a percentage of Total Spend. Above 40% is a structural problem requiring immediate action on match types and negative keywords — not just a budget adjustment.

Pre-calculated per entity by the pipeline's recovery model. It is the sum of recovery_potential across all classified waste entities in the 30-day window — not a flat percentage of wasted spend.

Wasted spend is increasing over the 30-day window. This usually indicates new broad match keywords were added, search volume shifted to lower-intent queries, or bid increases are attracting irrelevant traffic.

Waste is consistent but not growing. You have a structural gap — specific terms or match types are consistently draining budget — but no new damage is accumulating.

The campaigns with the largest blocks have the highest absolute waste. Fixing these delivers the most recovery in monetary terms, regardless of their percentage leakage rate.

Waste distributed across many small campaign/subcategory combinations indicates a structural match type problem — not isolated campaign issues. Broad match is triggering irrelevant queries account-wide.

A large spread means some campaigns are running efficiently while others are structurally broken. The problem is not account-wide — it is campaign-specific and highly actionable.

All campaigns are leaking at a similar rate. The waste is systemic — likely driven by account-wide match type settings, a universal audience issue, or a shared negative keyword gap.

A single click could be accidental or an outlier. The 2-click minimum ensures every term in this table has demonstrated a pattern of attracting spend without converting — not statistical noise.

The pipeline has high confidence this term will never convert based on multi-signal analysis. Adding it as an exact negative keyword will immediately stop the spend bleed.

The term is wasteful in the current window but has borderline signals — it may have converted in previous periods or has very low click volume. The pipeline needs more data before recommending exclusion.

The pipeline has insufficient data to classify this term confidently. It appeared in the zero-conversion list but doesn't yet meet the threshold for a negative recommendation.

NEAR_EXACT is Google's close variant matching on exact match keywords — triggering for plurals, misspellings, abbreviations, and implied words the algorithm considers equivalent. In accounts using predominantly exact match, NEAR_EXACT is frequently the largest waste category. Advertisers assume exact match is tightly controlled when Google's close variant logic is significantly broader than it used to be.

Broad match triggers for semantically related queries — including many with entirely different purchase intent. A high BROAD share means your account is spending on searches well outside your product category.

Phrase and near-phrase match triggers for queries containing your keyword phrase with additional modifying words. High waste here usually means the extra words are significantly changing purchase intent.

Wasted spend as a percentage of total spend for that device. A high waste_rate on a device means that device's traffic is converting poorly relative to how much it's spending — not just in absolute terms.

Mobile traffic frequently underconverts for B2B, high-consideration purchases, and desktop-optimised checkout flows. If mobile has the highest waste_rate, your landing page experience on mobile is likely the root cause.

Tablet traffic is the smallest device segment for most accounts but can have a disproportionate waste rate. Often indicates the landing page has a layout issue on tablet screen sizes.

A zero-conversion filter would flag every low-volume hour as waste. The dynamic CPA method is more accurate — it measures whether each hour's spend was justified by the conversions it actually produced, relative to your account's efficiency baseline.

Off-peak hours accumulate spend without conversions because purchase intent is low. The algorithm continues bidding because it doesn't daypart by default.

Hours with fewer than 50 clicks are excluded from the top wasted hours table. Below this threshold the waste calculation is statistically unreliable — a single outlier conversion or non-conversion would swing the result significantly.

Block this specific query. The waste pattern is specific enough that only this exact query needs to be excluded — blocking the phrase would be unnecessarily broad.

The waste pattern is broader than a single term — all queries containing this phrase should be blocked. Common for words like 'free', 'DIY', 'tutorial', or competitor names triggering irrelevant traffic.

A campaign qualifies for reallocation only if: it has at least one conversion, its CPA is below the account average CPA, AND it is losing more than 5% of impression share to budget. All three conditions must be met — this prevents the table from surfacing campaigns that convert but have no room to scale.

The estimated additional spend this campaign could absorb if budget constraints were removed, calculated from its lost impression share percentage. This is the maximum reallocation target — not a guaranteed conversion increase.

A steep CPC curve means improving Quality Score from low to medium levels will produce a disproportionate reduction in cost per click. The keywords in the 3–5 QS range are your highest-value improvement targets.

Your CPC is not highly sensitive to Quality Score in this account. Either your QS distribution is already healthy, or you operate in a market where Ad Rank is primarily bid-driven rather than QS-driven.

The algorithm is trapped in a learning loop — optimising for clicks that never buy. Over half of waste coming from EXPLORATION means campaign structure and budget pacing adjustments will outperform keyword pruning. The longer it runs without structural constraints, the more confidently it wastes budget on the wrong audience.

Waste is driven by weak RSA asset combinations — ads with low ad strength or poor headline/description variety that the algorithm keeps serving despite low conversion rates.

Search term intent mismatch is the primary driver — specific queries are repeatedly triggering your ads with zero purchase intent. This is the most directly actionable category.

You are overpaying at auction — winning clicks at a cost above what your conversion rate justifies. This is often caused by aggressive target CPA settings, broad match inflating competition, or low Quality Scores forcing higher bids to maintain position.

The number of distinct entities contributing waste in this category. A high entity_count with moderate total waste means the problem is widespread but shallow — many small fixes needed. A low entity_count with high total waste means a few specific campaigns or terms are responsible — higher-leverage, fewer fixes.

The 30% discount accounts for standard market volatility, the reality that not all paused spend will perfectly remap to incremental conversions, and the time it takes for campaign algorithms to adjust to structural changes. Real-world accounts typically recover between 55% and 85% of identified waste when fixes are applied.

Recovery is not money that appears in your bank account. It is budget currently spent on zero-return activity that, when stopped, becomes available to redeploy to efficient campaigns. Your total monthly spend stays the same — your conversion volume and ROAS improve.

The ROAS proxy uses a 3x average CPA assumption: daily_conversions × avg_cpa × 3 ÷ daily_cost. This is a conservative revenue estimate — actual ROAS will vary based on your product margins and actual conversion values.