Google Ads age-demographic audits for agencies: understand coverage, Unknown ages and actual inclusion settings

Define the client's age-review question

An age-demographic review should start with the client's actual advertising question. The agency may need to explain which reported groups contributed value, investigate a previously excluded range or compare acquisition costs within one campaign. Those questions call for different evidence. A generic account age table should not automatically become a recommendation to remove its smallest or least efficient row.

Record the requested period, metric and population before ranking groups. For commerce, the useful question may concern recorded value relative to spending. For lead generation, it may concern conversion acquisition cost while lead quality remains a separate business observation. The age range with the most spend is not automatically a defect.

The business objective matters as much as the reporting label. A client can sell the same offer to several adult age groups, and the person interacting with an ad is not necessarily the product's eventual user. The agency should not infer commercial eligibility or preference merely from a product description and an age bucket.

Define a useful deliverable. It might be a clearer explanation, a campaign-level investigation, retention of an intentional setup or preparation of an eligible supported setting change. The review can produce a sound decision without a live mutation. A provider-reported demographic pattern is evidence for analysis, not automatic authority to narrow reach.

Agree on what best means in the particular conversation. Most recorded value, highest return and lowest acquisition cost can identify different ranges. A client may already have supplied the intended metric in a preceding message, in which case the assistant should retain it. Where the choice remains unresolved, ask the one missing metric question before ranking. This keeps the analysis useful without forcing the client to restate the account, period and business brief.

Understand reported demographic categories

Google's demographic guidance describes both available demographic information and estimated categories. The agency should present reported age ranges as advertising segments rather than independently verified customer birth dates. The report does not provide a census of everyone who bought from the business.

The age label describes the selected source category. It is not an individual identity, exact age or proof of who used the product. Keep the statement proportional to the report: recorded metrics appeared under this returned range in the selected population.

Business customer data can answer a different question. A sales system may know an account holder's details or a booking participant's age, but that does not make its population identical to the Ads demographic report. A comparison needs deliberate definitions rather than a relabelled chart.

Use this boundary to improve the client discussion. The agency can identify a reported pattern and ask which campaign evidence could explain it. It should not describe an age group as inherently valuable or unvaluable without the actual metric and business context supporting that conclusion.

Read the current age-report collection

The current ppctor demographic helper requests the age-range view with the selected date segment. It retrieves the age criterion type, its negative flag and impressions, clicks, cost, conversions and conversion value. These fields supply both performance context and the declared included or excluded setting evidence.

The account path folds returned rows by age enum. The campaign path additionally groups by campaign identity. Neither compact collection is a complete list of every individual customer, and neither provides a full customer journey simply because it includes conversions.

The helper requests age, gender, income and parental views concurrently as separate queries. This separation follows their distinct provider resources. It does not turn their independent results into one jointly segmented demographic population.

The age finding is informational. It shows a metric table and readable summary without a scored fix. A local review status does not mean the account failed an age-quality test. The agency should explain the actual observations before choosing any later configuration workflow.

Keep the source identity attached to exports and follow-up explanations. The current view is built from demographic criteria and their reported metrics, not from a customer database or a survey. A PDF heading that simply says customer ages would broaden the claim. The agency should use the same accurate population description in Slack, a shared report and the handoff so presentation changes do not change the meaning of the evidence.

Preserve demographic-reportable coverage

The current account presentation explicitly names demographic-reportable campaigns as its scope. It labels the aggregate Reported total rather than claiming full account spend. Its caveat excludes Performance Max and App spending from this ad-group demographic collection.

Other campaign types contribute when the provider returns applicable demographic rows. Dimension and country eligibility can further limit the collection. The agency should not assume that every account metric is represented merely because several familiar channel names appear in the report description.

An account using substantial Performance Max activity can therefore have a demographic reported total below a separate account total. That difference is a scope issue to reconcile, not automatic evidence of missing money or incorrect arithmetic. Compare like populations before diagnosing the discrepancy.

The current age query has a date predicate but no explicit enabled-campaign filter, despite an older comment describing enabled campaigns. Follow the actual read when explaining coverage. Current status and historical reporting are separate facts; do not relabel the collection as only campaigns enabled today.

Reconcile coverage with base metrics before comparing efficiency. A demographic value total and cost total can each describe a narrower population than another account report. Comparing only their displayed ratios can hide that mismatch. The specialist should establish which activity belongs to both populations and what remains outside the demographic collection, then explain the actual requested comparison rather than treating two similar-looking percentages as proof of identical scope.

Separate current settings and period activity

The helper applies the requested date segment using the configured account time zone. State the actual interval when interpreting age performance. A completed rolling period and a calendar month are different questions, and a current partial day should not silently replace a completed-period comparison.

The negative criterion flag establishes the returned current setting classification. Metrics describe the selected reporting interval. An excluded row can therefore still have historical activity without proving that a current exclusion failed. The setting may have changed during or after the interval.

Conversely, zero impressions do not prove exclusion. A currently included group can have no activity in the selected population. Read the actual setting evidence and preserve the distinction between eligibility and observed delivery.

If the client asks why a range served during a past period, current configuration alone cannot reconstruct every historical setting. Establish relevant chronology where needed. The agency should not turn a present exclusion into a claim that the range was excluded for every date in the report.

Record when the setting was inspected as well as the reporting dates. A reviewer opening the same conversation later should know that the criterion flag was observed at preparation time, while the cost and conversions came from the selected interval. This simple distinction prevents a historical row from being treated as an execution error and prevents an old status observation from being presented as the current campaign setting.

Read the canonical age rows correctly

The current presentation shows six adult age ranges and an Unknown row in a consistent order. That readable structure helps compare reports, but the formatter can display a canonical row even when its key was absent from the returned collection.

For an absent key, the displayed metrics default to zero and its status becomes a dash. Treat this as a presentation fallback, not proof that the provider explicitly returned a measured zero record for that group. The underlying returned population determines what was actually observed.

The top aggregate uses actual returned rows. If an unexpected returned key appears, its metrics can affect that total while the canonical body does not give it a named row. A complete reconciliation should examine the actual key population instead of assuming the visible familiar ranges exhaust every returned value.

Do not manufacture exact individual ages inside a range. The current report offers category-level evidence. A request for a narrower age cutoff needs a genuinely supported provider definition and actual tool path, not interpolation from the compact group label.

Keep Unknown ages visible

The linked Google guidance explains that Unknown represents demographics the provider could not identify and that excluding it can narrow reach substantially. Keep that brief meaning attached to the actual row. Unknown is a legitimate reported category, not a synonym for bots or unusable traffic.

The current table includes Unknown metrics in its reported total. Removing the row from a client explanation can change the apparent coverage and distribution. A readable summary can emphasise known ranges while still naming how the unknown population was treated.

Unknown should not be assigned to known ranges by local guesswork. Its value and conversions do not tell the agency which identified ages would have received them. Dividing it proportionally across visible groups would create a fabricated distribution.

If the client explicitly considers excluding Unknown, establish the exact target scope and actual supported effect. A ranking that omits Unknown does not itself authorise a change to that category. Keep reporting selection and advertising reach as separate decisions.

An explanation can separate an unknown-only result from the inclusive aggregate without assigning identities to it. The owner can see how much selected reported cost and value belongs to that category and decide what question to investigate next. This is more useful than calling it leftover data. The category is part of the actual observed population, and its presence should remain visible whenever the client asks for a complete reported distribution.

Interpret Included, Excluded and Mixed

The helper observes the returned negative flags across contributing criteria. Included-only evidence produces Included, excluded-only evidence produces Excluded, and a combination produces Mixed. This is a settings aggregation rather than a status inferred from clicks or impressions.

Mixed can reflect different group configurations within a campaign or different configurations across the account. It does not identify which exact child should be edited. A follow-up mutation requires the relevant current resource and parent evidence rather than the aggregate word alone.

Included does not guarantee future delivery, and Excluded does not mean historical metrics must be empty. The classification describes the returned setting population. Performance and chronology still have their own evidence roles.

The current summary adds an exclusion note when a canonical group is purely Excluded. It does not use Mixed for that note. Absence of the note therefore does not establish that no contributing exclusions exist. Read the table status and relevant actual criteria before making a complete inclusion claim.

Mixed also does not establish a proportion. It means both flags occurred, not that half the budget or half the people were excluded. The compact helper records the presence of each state rather than a weighted share of eligibility. A request for the contributing groups needs actual group evidence. Avoid turning the word into a percentage, a delivery estimate or a ready-made list of resources to change.

Distinguish the reported total and summary

The reported total sums all returned base metrics, including Unknown. Its ratios are computed from the summed numerator and denominator rather than an unweighted average of group ratios. This is the correct population for interpreting that displayed aggregate.

The current metric-led summary deliberately filters Unknown before calculating its overall sentence and selecting the group with the most value or conversions. That population can differ from the reported total. The agency should not present the summary ratio as guaranteed identical to the inclusive table ratio.

Make the selected population explicit in a conversation. If the user asks for all reported age results, include Unknown in the aggregate. If it asks for identified ages only, preserve that narrower scope and state it. Do not silently change the denominator to obtain a stronger-looking headline.

A request for account-wide performance needs the proper account metric as well as coverage reconciliation. The demographic reported total does not become full account spend because its ratio was calculated correctly. Arithmetic correctness and source completeness are different requirements.

An invented arithmetic example makes the population difference clear. Suppose identified ages have 200 of cost and 600 of value, while Unknown has another 100 of cost and 100 of value. The identified-only ratio is 300%, but the inclusive ratio is about 233.33%. Neither calculation is contradictory when its population is stated. These numbers are illustrative arithmetic, not an observed client result or a recommendation to exclude the unknown category.

Interpret commerce value and ROAS

For an ecommerce-classified account, the current table presents conversion value and ROAS. ROAS divides recorded conversion value by cost and displays the ratio as a percentage. A ratio of four therefore appears as 400%, not a separate four-unit money figure.

The selected value is the standard Ads conversion-value metric. It does not automatically mean purchase-only sales, accounting turnover or profit. The actual contributing conversion definition needs to be established if the client asks what the value includes.

A group with no cost has no defined return on spending. The formatter shows a dash instead of inventing zero. A group with positive cost and no value can have zero recorded return for the selected evidence. Those conditions should remain distinct in ranking and explanation.

Confirm actual currency when discussing financial effects. The compact money presentation uses a dollar symbol, but that formatting alone is not proof of the client's account currency. An agency working across markets should preserve the real currency context before recommending a budget or target decision.

Interpret acquisition cost and lead quality

For the other account classification, the current table presents conversion rate and CPA. Conversion rate uses conversions divided by clicks, while CPA uses cost divided by conversions. These are defined from the selected recorded metrics, not from an independently verified list of unique qualified customers.

Fractional conversions can contribute to the reported totals. The display limits their precision while retaining the underlying calculation. Do not round every contribution into a unique person or invent a physical explanation for a fractional attributed result.

CPA is undefined when the conversion denominator is zero. That does not mean a group has zero acquisition cost. Keep the dash and explain the observed cost and missing conversions when the client needs that context.

Lead quality requires its own business evidence. A lower recorded CPA does not establish a higher close rate, better enquiries or more profitable customers. The agency can investigate those outcomes separately without renaming the conversion metric as a sales-quality measure.

Keep the denominator clear when comparing with website analytics. This compact conversion rate uses selected Ads conversions and clicks. A session-based website rate can have a different denominator and population. The client can request both, but they should be named accurately rather than presented as two equivalent measures of the same funnel step. A difference between them is a question to interpret before it becomes a tracking-defect conclusion.

Establish the conversion definition

The compact age table selects general conversions and conversion value. It does not separately enumerate every action name or category. A question about purchases, calls, leads or a named form action therefore needs its requested conversion scope established.

Keep the client's intended aggregate intact. An account can deliberately use several primary conversion actions with assigned values. The agency should not remove legitimate contributions or start a deduplication project merely to simplify an age comparison the client requested.

An individual Ads conversion compared with a specific Analytics event requires the corresponding named definitions and compatible evidence. A general age conversion total cannot be relabelled as that event because the counts look similar.

Separate a metric clarification from an account edit. Explaining which actions contribute does not authorise changing their roles. A supported conversion-role mutation remains an exact owner-approved decision with its own target, requested effect and result verification.

Assess scale before narrowing reach

A striking ratio from a small population can deserve attention without justifying a major reach restriction. Look at the cost, clicks and conversions behind the comparison. The current age finding does not provide a statistical certainty score or an automatic exclusion threshold.

Choose a reporting window relevant to the client's operating pattern. A short promotion can answer a different question from a steady acquisition period. Do not impose an arbitrary universal minimum and call it a provider requirement.

Investigate whether the pattern persists in the relevant campaign context. Account age rows can combine different offers and strategies. The row's label does not make those underlying populations identical or isolate age as the cause of a performance difference.

State the actual evidence and next decision. The agency can report that a selected range had a higher recorded CPA in the interval and assign a focused investigation. It should not claim that every person in that range is a poor customer or that excluding it guarantees better results.

Name the observation that could change the recommendation. It might be a particular campaign breakdown, the contributing group criteria or an established downstream lead-quality report. This is more useful than a generic request for more data and reduces unnecessary provider activity. The agency can continue collecting evidence toward a real decision instead of cycling through every available demographic view merely because the account has an uneven result.

Investigate actual campaign and group context

The account view aggregates contributing demographic-reportable campaigns. Brand search, prospecting and different offers can distribute differently across age groups. A follow-up should identify the campaigns relevant to the concern rather than apply one account ratio to every campaign.

Mixed inclusion status can require group-level inspection as well. A campaign may contain several groups with different criteria. The compact aggregate does not provide enough identity to choose one child for editing without the relevant current read.

Keep the same period and metric when narrowing the analysis unless the user changes the brief. A named-campaign follow-up needs the appropriate campaign population, not a relabelled account age total. Current setting eligibility can then be assessed separately.

Scope the proposed effect deliberately. One group can need a correction while the rest of the account remains appropriate. An account summary is useful context, but it does not authorise a silent blanket rewrite of demographic reach across every contributing campaign.

Do not fabricate combined demographic groups

Google's age-view reference states that combined age-and-gender segmentation is not supported by this API reporting approach. Independent age and gender results cannot establish the joint performance of one age-and-gender subgroup.

A total for one age range and a total for one gender can describe overlapping but unspecified populations. Multiplying shares or assigning one table's values to the other creates an invented joint distribution. The assistant should not output that as measured account evidence.

The client can still receive separate useful views. Explain which dimension each report describes and retain its scope. A requested combination needs a genuinely available compatible source rather than a locally manufactured answer.

The same discipline applies to other combinations. A named campaign filter and an age dimension can be supported by actual reporting evidence, while an arbitrary compound audience claim may not be. Establish the specific requested composition instead of assuming every set of labels can be joined.

Two separately available charts can still answer a useful broader question when their independence is explicit. The client may want to understand age and gender patterns side by side. Provide those individual observations without inventing their intersection. The assistant should explain the concrete unavailable composition and offer the actual supported presentation, rather than falsely claim that no demographic information exists or output a locally estimated joint table as measured performance.

Distinguish unread, empty and unsupported data

The account age card can show a loading placeholder while its deferred report phase is pending. Once the report phase completes, unavailable evidence has a separate failure mapping. Loading should not become a permanent explanation for work that already failed.

The current demographic helper uses one concurrent collection for four views. If one view throws, the combined helper rejects and the account wrapper substitutes empty dimensions while recording demographic failure. Do not promise that a successful age result is always preserved when another demographic dimension fails.

A successful empty dimension can hide the informational card. That is different from a recorded unavailable read. Campaign applicability also matters; the current campaign presentation explicitly omits Performance Max and handles eligible channel failures separately from unsupported populations.

The agency should identify the actual condition before drawing a conclusion. Unread data is not zero performance, and an unsupported report is not a clean targeting certificate. Obtain the missing relevant observation before presenting a complete age distribution or preparing a live setting decision.

Follow an invented agency age review

This case is invented. Pine Harbour Media manages Cedar Court Classes. The client asks which age range delivered the best recorded return in a completed period. The agency confirms the conversion-value definition and notes that the demographic collection does not include the account's Performance Max spending.

The specialist presents the reported age population with Unknown retained in the inclusive total. It separately identifies the known-age ranking requested by the client, without claiming the compact summary's known-age ratio is the same as the inclusive reported total.

One range has Mixed status. The agency does not infer that it is half excluded or choose a random group to edit. A focused follow-up establishes the relevant Local Classes campaign and its actual contributing group configuration.

Cedar Court Classes determines that the immediate task is understanding a deliberate existing group difference. No live mutation is prepared merely because the informational report contains an uneven distribution. The agency records the business reason and leaves a separate performance investigation where evidence is still needed.

If the owner later requests a supported demographic change, the assistant must establish exact current target identity, provider eligibility and the actual preparation path. The proposal names the real group and campaign, and the live effect follows owner confirmation.

The handoff preserves the reportable population, Unknown treatment, conversion definition, Mixed-status investigation and actual decision. The invented example illustrates accountable interpretation rather than a promised uplift or automatic demographic editor.

Pine Harbour Media also records what did not become an account action: the owner requested an explanation and retained the deliberate group difference. When another colleague later reads the thread, the status table does not imply a new exclusion was executed. If Cedar Court Classes asks for a different campaign review, the specialist carries forward the metric brief while obtaining that campaign's relevant evidence. Continuity reduces repetition without turning the first campaign's observed criteria into facts about every other campaign.

Choose a supported response deliberately

A review can lead to retention, another reporting read, a campaign-context investigation or an eligible supported edit. Select the response that follows the evidence and user instruction. A convenient mutation should not decide the strategy merely because it is available.

The current convenience catalogue has no dedicated universal age-setting action, and the informational age brick has no direct fix. A generic schema-mediated edit must establish the supported resource, requested fields, current target and applicable result checks before it can be described as available.

Creating an audience signal with demographic ingredients is a different action from changing an existing age exclusion. Do not treat those tools as substitutes for one another. The prepared effect needs to match what the client actually requested.

Creative campaign planning can remain conversation when the supplied brief is sufficient. Suggesting a test message or audience hypothesis does not require a provider call solely to draft text. Claims about existing account criteria and measured outcomes still require relevant evidence.

Present exact owner-approved changes

A live proposal should show the exact target, real parent group or campaign relationship and requested action. The owner should understand the resulting reach effect without reading internal criterion IDs or a technical mutation body.

Show each logical action as a visible step in the supported confirmation. When a larger request cannot be reviewed safely as one proposal, ask for smaller requests before execution. Never apply a silent partial interpretation. A reporting summary and a later setting change remain distinct deliverables, even when they concern the same demographic population.

The pending message states that no changes have been made yet and the proposal expires in fifteen minutes. The connected owner confirms before live execution. A descriptive age comparison or strategic explanation does not itself authorise a mutation.

Current validation checks applicability before the write, while readback verifies the declared resulting state afterward. Neither should be represented as a performance guarantee. The proposal must describe the actual supported effect even if the original conversation used broad optimisation language.

Verify the requested setting result

Successful configuration readback establishes the declared requested fields and relationships. It does not prove that the campaign immediately achieved a better ROAS or CPA. Keep the executed-and-verified receipt clear while retaining the separate business-outcome review.

A visible change-in-progress state acknowledges confirmation while provider work continues. The owner should not have to repeat a click because the result takes several seconds. A delayed Slack receipt is different from a rejected provider request.

If an edit was sent but full verification is unavailable, reconcile the exact current target and durable execution record before another attempt. The setting may already have changed. Preserve separate outcomes where a multi-action request has verified and unresolved parts.

A later restoration request is a new supported edit with current identity and exact intended result. Do not promise universal rollback from a vague request to undo the age review. Historical metrics and an old report are context, not permission for blind replay.

Present readable age tables and charts

A table is useful when the client needs age metrics and inclusion status together. A daily chart answers a different question and requires daily evidence. The assistant should not distribute period totals across dates or manufacture a trend from the canonical age rows.

State whether Unknown is included in the requested chart or aggregate. A known-age-only comparison can be useful, but the population should be visible. Do not silently drop a major reported category merely to make a chart look cleaner.

Native Slack charts support a readable maximum of three selected metrics or segments and twenty points per series. Longer time periods should use a stated coarser breakdown when daily detail was not required. An explicitly daily request beyond that limit needs the appropriate explained split.

Keep ordinary metrics to at most two displayed decimal places and current-year dates concise. Names and actual effects are more useful than provider enums. Preserve undefined ratios rather than padding them into apparently measured numbers.

Keep the inclusion status beside the performance evidence when the question concerns exclusions. A chart of cost alone cannot establish whether a range is currently included. The owner may need a short status table alongside the chart, using the exact same declared scope. That presentation answers both parts of the request without forcing a setting inference from the bar height or treating a missing bar as an excluded criterion.

Continue the age discussion in Slack

A focused opening can ask for age performance over a selected interval and metric. The assistant should use established account context where the intended metric is clear, then state its actual definition. Genuine ambiguity needs a narrow question rather than a restart of the whole brief.

A follow-up about one campaign should preserve the prior period unless the user changes it, but needs that campaign's relevant evidence. Account context can guide the read without becoming proof of a new campaign claim.

Explaining Mixed status or Unknown does not require a fresh provider call when the necessary conceptual and returned evidence is already available. Current setting changes and new reporting populations need their appropriate reads. This avoids redundant activity while preserving correctness.

Keep user mentions before the final feedback text and controls. Progress messages should be visible without a user mention. The ordinary reply should name the observation, concrete next decision and known result rather than expose raw IDs, technical payloads or undefined parents.

Leave an accountable agency handoff

Record the business question, selected period, reportable population and conversion definition. Include Unknown treatment and the exact status interpretation. Another specialist should know whether a conclusion concerns absolute volume, efficiency or a current criterion setting.

Preserve the actual campaign or group investigation that changed the recommendation. Mixed status should remain a scoped configuration question until the relevant children are known. A completed explanation should not be recorded as a live edit that never occurred.

Retain deliberate exceptions with their owner and reason. Future reviewers should not repeatedly propose age exclusions because a table remains uneven. The appropriate decision follows the client's objective and evidence, not an expectation that every demographic row must have the same performance.

Close the handoff with one concrete continuation point. State whether the task is finished as an explanation, waiting for a named observation or completed as a verified supported edit. This gives the next operator an actionable record and keeps operational logs faithful to actual execution. A conversation request and an account mutation are different events, even when they concern the same age range.

Use the agency audit workflow for account context, the device-performance review for another dimension and the approval workflow for precise supported edits. Each retains its own population and evidence boundary.

Answer common age-review questions

Is reported age an independently verified exact age?

No. The table uses provider demographic categories.

Does Reported total equal full account spend?

No. The current demographic collection has channel and other coverage limits.

Does zero activity prove a group is excluded?

No. Inclusion status comes from returned criterion flags.

What does Mixed mean?

Both included and excluded criteria contributed to the aggregated group.

Is Unknown included in the table total?

Yes. The current table includes it, while the metric-led summary filters it out.

Can age and gender totals establish a combined subgroup?

No. Independent marginal reports do not provide that joint performance.

Does the age brick have a direct automatic fix?

No. A live edit needs an actual supported preparation path.

Does verified execution guarantee better return?

No. It establishes the requested configuration; outcomes need separate evidence.