How to Read Your AI Max Signal Report

A composite score measuring how closely Google Ads conversion numbers match GA4 ground truth. Calculated from signal health intelligence across all campaigns. Below 60 means the bidding algorithm is making decisions on unreliable data.

The average percentage difference between what Google Ads reports as conversions and what GA4 independently confirms. A 30% drift means for every 10 conversions Ads claims, GA4 only sees 7.

The percentage of all_conversions that are NOT click-attributed. This includes view-through, cross-device, and algorithmically modeled conversions. High modeled noise means the algorithm is optimizing on estimates, not observed behavior.

Average readiness score across all campaigns from the maturity model. Measures how close campaigns are to the exploitation phase where they can scale efficiently. Low readiness + high spend = paying for learning.

Signal quality and attribution accuracy are aligning. The bidding algorithm is receiving increasingly reliable data. Continue current tracking improvements.

The gap between what Ads reports and what GA4 confirms is widening. This compounds daily — each bid decision uses worse data than the day before.

The user clicked your ad before converting. This is the strongest attribution signal — a deliberate user action. The larger this slice, the more reliable your performance data.

The user saw your ad but didn't click, then converted through another path. Google counts these by default. If this exceeds 30% of total conversions, the algorithm is optimizing on impression exposure, not purchase intent.

Conversions that Google infers using machine learning models. These are legitimate statistical estimates, but they introduce noise the algorithm cannot distinguish from observed behavior.

Ads and GA4 broadly agree. Bidding algorithm has clean signal. No action needed.

Meaningful discrepancy. Common causes: duplicate conversion actions, mismatched attribution windows, cross-device gaps.

The algorithm is bidding on phantom conversions. This inflates CPCs account-wide and degrades ROAS across all campaigns. Every day this persists compounds the damage.

More than half this campaign's conversions are algorithmically estimated. The bidding algorithm cannot distinguish real demand from statistical inference at this noise level.

This campaign's conversion data is predominantly click-attributed. The algorithm has reliable signal to optimize against.

The share of conversions for this action that come from view-through, cross-device, or modeled attribution rather than direct clicks. Higher = less reliable.

Where the conversion tag fires from (website, app, import, etc). Tags firing from unexpected sources may indicate misconfiguration or duplicate firing.

Over 30% of this campaign's conversions are cross-device modeled. The algorithm may be over-crediting mobile impressions for desktop purchases it can't directly verify.

Most conversions happen on the same device as the click. Strong single-device attribution signal.

The algorithm has enough conversion data to bid confidently. Budget increases in this stage produce predictable returns.

The algorithm is actively testing new audiences and inventory. Some exploration is healthy, but excessive exploration means the algorithm lacks guardrails.

Performance is actively declining — usually from creative fatigue, audience saturation, or major account structure changes.

The algorithm is forced to use low-performing assets because it has no better options. This drives up CPAs and pushes spend toward junk inventory.

This audience segment consumed budget with zero conversions. The algorithm will continue targeting it unless you add exclusions.

Not enough conversion data to classify. The algorithm is still exploring this segment.

This campaign has accumulated enough conversion data for the algorithm to bid confidently. Budget increases will produce predictable returns.

The campaign is still building its conversion base. Cutting budget now wastes the learning investment already made.

This tag is producing more modeled conversions than click-attributed ones. The current implementation is fundamentally broken — patching won't work.

This action may be firing multiple times per session or counting the same conversion from both Ads and GA4.

This campaign has no reliable conversion signal AND is still in the expensive exploration phase. Every dollar spent is both wasted and teaching the algorithm wrong patterns.

This campaign's performance is declining. The algorithm has lost signal quality — usually from creative fatigue or audience saturation.

Signal quality is getting worse despite interventions. This usually means the root cause (duplicate tags, wrong attribution model) hasn't been addressed.

This conversion source is producing mostly modeled data. The tag is either misfiring, firing on the wrong event, or duplicated across multiple containers.

This source's conversions are predominantly click-attributed. The tag implementation is correct.

These combinations are getting clicks (spend) but not conversions (high CPA). The AI headline is attracting the wrong audience or misrepresenting the landing page.

The AI is sending significant traffic to a page that underperforms. The page itself may have UX issues, or the AI is matching the wrong audience to this page.

The algorithm cannot hit this target with current signal quality and creative assets. Either the target is unrealistic or the campaign needs structural fixes first.

The target is set so low that the algorithm cannot find enough qualified traffic. You're leaving conversions on the table.

More than half this campaign's conversions are algorithmically estimated. The bidding algorithm is effectively guessing where to allocate your budget.

The algorithm's expansion into broad and auto match is costing more per conversion than your manually controlled keywords. The AI is finding volume, but at a premium that may not be worth paying.

The algorithm is finding conversions as efficiently as or better than your manual keywords. This is the signal to trust the expansion and potentially increase budget.

Your hand-written ad copy generates better returns than Google's automatically created assets. The AI copy may be generic, off-brand, or mismatched to landing page content.

The algorithm is expanding into search terms that never convert. This is a match type and negative keyword problem.

The algorithm is spending on audience/placement testing that isn't yielding results. This is a targeting guardrails problem.

These are the highest-impact waste sources that can be fixed in minutes. Pausing a single bad search term or excluding a wasteful placement recovers budget immediately.

The AI is almost exclusively capturing people already searching for your brand or exact products. It's efficient but not finding new customers — you're paying for demand you already own.

This campaign's CPA is higher than average, but it's acquiring genuinely new customers with proven lifetime value. The CPA premium is an investment, not waste.

This campaign is paying full acquisition prices for users who aren't delivering meaningful lifetime value. It may be retargeting at full price rather than prospecting.

Daily spend swings wildly. The algorithm is alternating between aggressive bidding and pulling back, which makes performance unpredictable and planning impossible.

Markets where CPA, conversion rate, or location score deviate significantly from the account average. The recommendation column tells you whether to increase targeting, monitor, or gather more data.

Hours of day where the algorithm's spend concentration doesn't match conversion probability. The recommendation tells you whether to adjust ad schedules.