Review parental-status evidence without guessing who a purchase is for
Google Ads parental-status audits for agencies: understand reported outcomes before narrowing campaign reach
Separate demographic labels, buyer intent and current targeting. Review real report coverage, reconcile Unknown data and explain supported next steps to the client.
Start with the client question
A parental-status review should begin with the decision the client needs to make. The agency might be explaining a demographic split, checking an inherited restriction or assessing whether an offer reaches the intended buyers. Those tasks require different evidence and should not automatically end in the same targeting recommendation.
Establish the selected campaign or account population, reporting dates and success measure. Most recorded value, strongest return and lowest acquisition cost can identify different categories. The assistant should use the already agreed objective where the conversation supplies it, then ask a focused question only when the remaining ambiguity affects the answer.
A product aimed at families does not establish that every purchaser belongs to the Parent reporting category. The commercial brief identifies the business objective; the demographic table describes the source's observed categories. The agency needs both without pretending one is a complete explanation of the other.
Define the deliverable before expanding the analysis. A readable explanation, a campaign investigation, retention of an intentional setting or a supported edit proposal can each be useful. A table with uneven results is not itself proof that the campaign needs a demographic restriction.
Record the question in language the next account manager can understand. Investigate why the existing seasonal campaign excludes one category is a concrete brief. Optimise demographics is too broad to identify the necessary evidence, owner decision or actual effect that would count as completing the review.
Explain the source labels accurately
Google's demographic guidance describes parental-status categories and demographic estimates. Parent, Not a parent and Unknown should be presented as source labels. They are not independently verified descriptions of every customer or a complete record of household relationships.
The API's parental-status enum distinguishes parent, not a parent and undetermined values, with additional unspecified and version-unknown values. The application needs to preserve the returned meaning rather than assigning an unidentified interaction to a known category.
The current reader does not provide the number or ages of someone's children, household responsibilities or the intended recipient of a purchase. Do not add these details to a report simply because they appear plausible for the business. A category heading is not evidence for a more specific personal profile.
Keep the language neutral and limited to the observation. Recorded conversions appeared in a category is a factual statement about the reviewed report. Parents prefer this product is a broader claim that requires evidence the demographic aggregate does not supply.
A client survey or customer database may describe another population with different definitions. It can inform a separate analysis when available, but similar headings do not make its records equivalent to the Ads report. Establish the source and scope before using it to explain an apparent difference.
Separate the buyer from the recipient
People can purchase products for someone else. A family-oriented catalogue may receive orders from relatives, friends or institutional buyers as well as parents. These possibilities explain why a product's intended audience is not identical to the purchaser's reported category; they do not establish any individual customer's identity.
The same distinction applies to enquiries. A person researching a service can act on behalf of another household member. The parental-status row does not expose that relationship. The agency should not convert a form submission's demographic category into a verified description of who ultimately receives the service.
Ask the client which buyer relationships matter commercially when that information affects the recommendation. A gifting business may want to retain broad reach, while another brief prioritises a particular campaign objective. The advertising data cannot choose that business preference on the client's behalf.
Do not treat Not a parent as irrelevant traffic merely because the product serves children or families. Its recorded value remains part of the measured population. Any proposal to restrict that reach needs actual campaign evidence and a clear intended effect, not a conclusion drawn from the label alone.
Creative can address the product use case without claiming to know a person's household. A practical agency recommendation might clarify gifting, availability or service benefits. That is different from using an aggregate category to make a personal assertion about everyone who sees the ad.
Understand the parental-status collection
Google's parental-status view identifies a resource associated with an ad group and criterion. The current ppctor reader requests parental-status type, the negative flag and aggregate impressions, clicks, cost, conversions and conversion value. It does not retrieve identifiable family records.
The account collection aggregates rows by returned enum. The campaign collection keeps campaign identity before folding the categories. An account observation can guide a campaign investigation, but the agency cannot relabel its account totals as the results of a particular campaign without obtaining that narrower evidence.
Parental status is one of four separately requested demographic dimensions. Age, gender and household income are collected alongside it through concurrent queries. That collection structure does not produce a joint demographic distribution or establish the overlap between a parent category and a particular income band.
The resulting finding is informational, with hidden status and no direct universal fix. It should not be sold to the client as a scored targeting defect. The useful material is the actual reported performance, retrieved configuration indications and the decision those observations can support.
Distinguish the fields received from the conclusion written by the agency. Cost and conversion value are source metrics; a judgment about campaign suitability is an interpretation. A clear report shows why the interpretation follows from the evidence instead of presenting it as another fact directly supplied by the API.
Keep campaign coverage explicit
The account demographic card uses Reported total for its aggregate. Its stated ad-group reporting population excludes Performance Max and App spending. A populated parental-status table is therefore not a full-account census of every advertising interaction or every conversion by default.
Dimension applicability narrows coverage further. Google's current guidance lists parental status for Display, Video and Demand Gen; it does not list the same option in its Search demographic choices. Do not infer parental-status evidence for every channel merely because other demographic dimensions return rows.
The reader still depends on what the provider returns for the relevant request. The presence of several eligible channel names in a caveat is not proof that every campaign contributed. Explain actual report coverage before comparing its cost with a separate account-wide spending total.
The current account query contains a date predicate but no explicit enabled-campaign restriction. Historical performance and present campaign status therefore remain separate facts. An old comment about enabled campaigns should not be repeated as the actual query boundary when the query does not implement it.
Preserve the scope in every output. A chart, spreadsheet and Slack sentence can all be accurate individually yet misleading together if one says reportable campaigns and another says entire account. The agency should carry the same population definition through the final conclusion.
Separate historical outcomes and current eligibility
The selected date range tells the agency when the measured interactions occurred. The retrieved negative flag describes configuration evidence. A currently excluded category can still have historical activity. Those observations need not contradict one another, and the assistant should not invent a timing explanation to reconcile them.
To connect an exclusion with a performance change, obtain actual change evidence. A decline in conversions does not prove who modified the setting or when it changed. Campaign demand, offers and measurement can also vary; the report alone cannot establish the causal sequence.
Align dates across supporting reports when the decision requires a direct comparison. A recent parental-status table and a previous season's product report answer different questions. They can both be useful context, but their differences should remain explicit rather than disappearing under a single performance heading.
Recent recorded outcomes can continue developing after an interaction date. A completed reporting window is not a guarantee that all conversions are permanently settled. The agency should use an appropriate interpretation of measurement maturity when a small difference drives a proposed eligibility change.
A takeover handoff should record the current setting and historical window separately. That lets another analyst reproduce the observation without assuming the retrieved exclusion existed throughout the whole period. Current configuration is a baseline, not a complete history of the account.
Retain unidentified results
Unknown describes information that has not been identified in the reporting category. It does not mean Not a parent, uninterested visitor or invalid lead. The agency should retain unidentified outcomes rather than assigning them to whichever known category makes the commercial story appear simpler.
Unidentified rows can carry real cost and recorded value. Their inclusion changes aggregate return and acquisition cost. A known-only comparison can be useful, but it needs that subset label; it cannot replace the inclusive Reported total merely because it produces a more attractive headline.
Excluding Unknown concerns reach eligibility rather than repairing a missing demographic label. It would not identify the remaining people's households or improve conversion definitions automatically. A supported proposal requires a reason that follows from the client brief and the actual campaign evidence.
When Unknown is large, explain the observed information boundary. The report shows an unidentified contribution within its population. It does not provide a household-level investigation capable of identifying every reason that an interaction remained unclassified.
Keep the category and its measured outcomes visible in a client export. Otherwise a reader outside the application can mistake known-band efficiency for the whole report. The document should preserve enough scope to make the result understandable even when the surrounding conversation is no longer available.
Check the Unknown mapping and retained totals
The canonical parental-status order uses UNDETERMINED for the displayed Unknown row. Google's enum also contains UNKNOWN, a return value for a value unrecognised by that API version. These values have different meanings. The reader preserves returned keys, while the table maps the ordinary undetermined classification to its human-readable label.
When the returned data includes UNDETERMINED, its metrics and criterion-derived state appear in the Unknown row and contribute to the reported total. An unexpected additional key still needs investigation: the total includes retained rows, while the visible category list follows its recognised classifications. The agency should reconcile any such difference before describing a complete breakdown.
The identified-group summary intentionally omits the visible Unknown category. Its known-category selection is not an inventory of every actual returned key, so it describes a different population from the total. An unexpected mismatch should be investigated explicitly rather than solved by adding an invented balancing number to the visible table.
Do not certify a full Unknown result from the heading alone. Check which enum keys actually arrived and which ones the displayed row represents. The report can remain useful for known observations while that reconciliation is performed, but its completeness claim needs to match the verified evidence.
This article does not describe that limitation as already repaired. The page documents the current reader and renderer rather than silently changing an unrelated runtime feature. A future verified mapping correction should update the implementation evidence and this explanation together so the agency can rely on the new behaviour.
Interpret configuration flags and Mixed status
The aggregation derives Included, Excluded and Mixed from returned negative flags. Both positive and negative indications produce Mixed. This describes the retrieved configurations across rows; it does not quantify the proportion of traffic excluded or establish that half the category's users are eligible.
An aggregated status does not identify every child target. Obtain the required campaign and ad-group context before preparing a change to a particular criterion. The agency should not guess which parent owns a restriction based on a category's account-level performance.
Impressions and eligibility indicators remain separate evidence. Historical impressions do not automatically invalidate a present exclusion, while no impressions do not establish that the category was excluded. Use the actual retrieved flags instead of deriving current settings from performance volume.
The current summary's exclusion note is triggered by purely excluded canonical rows. Mixed rows do not trigger that same note. Its absence therefore cannot prove that every underlying criterion is included; the agency needs to inspect the actual status and relevant scope.
A Mixed result can be intentional across campaigns with different briefs. Investigate the commercial rationale before treating it as inconsistent configuration. A supported correction should address the actual target and reason, not force every campaign into identical settings because the aggregate table contains one mixed label.
Reconcile total contribution and summary leaders
The current aggregate total sums actual returned rows. The highlighted summary works from canonical categories and excludes the displayed Unknown label. Those features do not necessarily describe the same population, especially where an undetermined key is retained outside the canonical presentation.
The summary selects a leader by absolute conversion value or conversion count, according to the account presentation. It does not select the highest return or lowest acquisition cost. Translate that highlight accurately instead of calling the largest contribution the best-performing category without specifying the metric.
If the client asks for efficiency, inspect the appropriate ratio and its denominator. If it asks for volume, compare volume. Both comparisons are valid within a verified population, but the summary intended for one cannot automatically answer the other.
Reconcile the underlying returned rows when totals appear inconsistent with visible categories. Fallback cells and unexpected keys can explain a presentation gap. Do not make up missing measurements or silently drop returned evidence to make the displayed sum resemble a preferred headline.
A useful conclusion states the observation precisely: a category supplied the most recorded value among the reviewed identified rows. It does not assert that the category represents the client's largest customer population, because the current report does not provide that customer census.
Use the right denominators
Total return is summed recorded value divided by summed cost for the selected population. Averaging the category percentages gives unequal spending rows equal influence and can produce a misleading total. Use the underlying sums when calculating the account or campaign result.
Acquisition cost is total cost divided by total recorded conversions. A category with no conversions has no valid finite acquisition-cost result in this calculation. A dash is different from zero, and the client should not interpret it as acquiring customers at no cost.
Recorded conversions can be fractional. Keep that precision in calculations while formatting readable outputs. Rounding every row into whole events before computing total cost per conversion can distort the answer, especially when the relevant comparison involves small numbers.
The current audit uses percentage presentation for return. A ratio of four appears as 400%. If a narrative uses a multiple instead, label it clearly. Consistent units prevent the client from mistaking two representations of the same value for contradictory measurements.
Currency also needs a verified account context. A compact dollar sign in a table is not independent evidence of which currency the account uses. Keep the true currency attached to any commercial recommendation and avoid unsupported comparisons with another market's amounts.
Evaluate commerce without inventing customer roles
Recorded conversion value follows the account's measurement configuration. It is not automatically profit, settled cash or a complete order ledger. Parental-status reporting inherits that definition rather than supplying a separate financial verification of who bought or consumed the product.
A high-value row can reflect the product mix, campaign mix or offer period. It does not establish that one household category is inherently more profitable. The current reader does not contain every margin, refund and fulfilment detail required to reach that broader conclusion.
For family-oriented products, distinguish the purchaser relationship from the product use case. A gift order can be commercially valuable even where the source category differs from the campaign's intended recipient. Do not infer a lost sale or unsuitable order from that demographic difference alone.
Ask for the client's relevant commercial facts when they affect the decision. Product priorities, customer acquisition goals and margin constraints belong to the business brief. The agency can incorporate supplied evidence without pretending those details were returned by the demographic API.
The immediate outcome may be to retain reach and review product-level evidence. That is a practical recommendation when the table supports an observed distribution but not a defensible exclusion. It avoids turning a narrow source label into a customer-value judgment the report cannot substantiate.
Evaluate recorded leads and downstream evidence
Lead-generation presentation commonly uses conversion rate and acquisition cost. These measure recorded actions in the selected population. They do not verify family eligibility, household needs or eventual customer quality merely because the row is labelled Parent or Not a parent.
Confirm what the client considers a lead. A submitted enquiry, phone call and imported offline outcome can represent different stages. The agency should not describe the aggregated count as unique qualified households unless the actual measurement definition supports that description.
A CRM comparison can add commercial context when the appropriate evidence is supplied. Keep its population and matching limitations visible. An aggregate parental-status report cannot assign a verified household role to each CRM record simply because both systems report a similar number of outcomes.
The lowest recorded lead cost may not identify the best commercial segment. Different enquiry mixes and downstream results can change the client's preference. Obtain the relevant business evidence rather than inventing a quality score from the demographic category or acquisition-cost ranking.
A clear interim answer gives the measured lead comparison, explains the chosen scope and identifies the specific missing fact. It should not require a new full-account audit if the only unresolved issue is the client's downstream definition. Keep the investigation focused on the decision.
Confirm the conversion definition
The reader requests conversions and conversion value. These fields do not independently prove a purchase-only total, a unique-customer count or a count of different families. Use the actual reporting semantics and account configuration before assigning a narrower commercial label.
A request about a named conversion needs evidence for that action. Do not estimate its parental-status distribution by allocating the aggregate total according to another row's share. Similar account proportions cannot replace the compatible segmented observation required for the specific question.
For an aggregate primary-conversion comparison, preserve the selected primary and campaign-goal context. Lead categories remain leads unless the client requests a particular action. Primary calls and offline contributions can also carry recorded value; they are not automatically excluded from every commercial interpretation.
Counting rules and duplicate implementation belong to a separate measurement review. The parental-status table does not diagnose which business outcome was recorded more than once. Do not subtract contributions locally in an attempt to repair a presumed deduplication problem the source evidence has not established.
Carry the chosen definition through the report. Calling the table conversions, the summary purchases and the recommendation customers can create three different claims from one field. The agency should keep the language stable unless another verified source supports a deliberate distinction.
Investigate campaign and offer differences
Account-level parental-status performance combines the campaigns that return applicable evidence. Different campaigns can promote different products, use different creative and pursue different conversion goals. An aggregated category difference therefore does not isolate parental status as the cause of weaker performance.
A campaign investigation can reveal that a seasonal offer drove the change while an evergreen offer remained stable. Obtain actual scoped evidence before making that statement. The account total alone cannot identify the contributing campaign or explain why its category distribution shifted.
Product availability can alter the commercial context. A promoted service or item may have become unavailable, or a landing page may have changed. These are investigation candidates, not facts established by the parental-status table. Verify the relevant event before attaching it to the performance explanation.
Budgets and campaign priorities can also change the population mix. A category receiving more recorded value may reflect increased exposure to a strong campaign rather than a universal preference. Preserve that distinction when the client asks whether the category itself improved.
Choose targeted additional reads that resolve the question. Recollecting every account surface creates work without improving certainty when the missing evidence concerns one campaign's offer or setting. A scoped review remains thorough when it obtains the facts required for the actual decision.
Account for the business calendar
Family-related buying and service demand can vary around the business's own calendar. Promotions, term dates and gifting periods can change campaign results. The agency should use verified client context rather than assuming a universal seasonal pattern from the parental-status label.
Compare periods that answer the question. A recent promotion and an ordinary trading window may not be equivalent tests of the same offer. A useful explanation names the difference and avoids presenting it as proof that the source category became intrinsically more or less valuable.
Keep date boundaries consistent across the report. The campaign cost window, conversion evidence and supporting commercial notes should not silently refer to different months. When a longer comparison is useful, make its purpose explicit rather than mixing dates under one headline.
Recent outcomes may need time to develop, particularly where the business closes leads later. The table cannot guarantee a settled commercial result simply because the interaction date has passed. Use the relevant measurement interpretation before responding to a small recent category difference.
Record the next review trigger in practical terms. A post-promotion evaluation or a review after a confirmed offer change gives the agency a useful follow-up. A vague instruction to watch demographics does not identify what evidence would justify revisiting the original decision.
Choose an evidence-based response
Evaluate the amount of measured activity alongside the ratio. A handful of interactions can create an extreme result that shifts with one later conversion. The current audit reports observations; it does not calculate a universal statistical guarantee that every category ranking will persist.
Attach cost exposure to the interpretation. A weak category that spent little has a different practical significance from a weak category absorbing a large part of the report. The appropriate response can be an explanation, further investigation or a supported proposal depending on the actual evidence.
Do not manufacture a minimum sample rule the application never applied. The agency can explain why a decision needs more evidence without pretending the table passed a formal test. If a separate statistical analysis is requested, define and perform that analysis explicitly.
Retaining an existing setting can be a justified outcome. The review might show intentional reach, a commercial rationale or a performance difference explained by campaign mix. Record that reasoning so another analyst understands why a tempting exclusion was not recommended.
A supported intervention should identify its intended effect and later review condition. Technical execution verifies the setting change, while subsequent performance provides another observation. The client should not be promised an uplift merely because the preparation stage found a valid editable field.
Avoid unsupported combined demographics
The current reader collects parental status independently from age, gender and income. Those separate totals are marginal distributions. They cannot establish performance for parents in one age band with a particular income label, because the joint population was not returned.
Do not multiply percentages to manufacture that intersection. Doing so creates a result based on an unstated independence assumption, not the campaign evidence. A neat chart can still be false when its source values were assembled from incompatible aggregate dimensions.
The current table also does not provide a complete product-by-parental-status or query-by-parental-status breakdown. A more detailed request needs a compatible supported source. Explain the precise unavailable scope instead of presenting a guessed allocation as observed performance.
Audience settings containing parental-status ingredients are configuration evidence, not a replacement for the missing joint report. An attached audience does not establish that every subsequent recorded interaction belongs to its labels or explain the whole historical distribution.
A helpful assistant can continue with the supported evidence. It can show the separate views with their actual definitions or identify the next compatible investigation. It should preserve the original request while making clear which combined claim the available collection cannot support.
Distinguish unavailable and empty reports
The four demographic queries share a Promise.all failure boundary. One rejected view rejects the whole collection, and the outer loader records the demographic section as failed. Do not claim that a successfully returned parental-status view remains independently available through that failure in the current implementation.
Loading, unsuccessful retrieval and a successful empty response are different states. None should be described as a completed finding that every category recorded zero outcomes. The agency needs the actual source state before explaining why a expected table is absent.
Canonical fallback rows provide a stable presentation when a key is absent. They are not measured observations. That boundary is especially important here because the current Unknown mapping can differ from the undetermined key actually returned by the provider.
A read retry can help with a temporary provider failure, while a compatibility or unsupported-population problem needs another explanation. Do not keep repeating an unchanged request as though every absence must eventually become data. Identify the actual failure or coverage condition where evidence permits.
Resume the original client question after a successful recovery. Preserve the campaign, period and measure rather than asking the client to restate the entire brief. Reporting recovery should resolve the missing evidence without discarding the context that already established the task.
Follow an invented agency review
Consider the invented agency Birch Harbour Media and its invented client Cedar Court Learning. The client asks whether a family-course campaign should exclude Not a parent after seeing a weak recent result. These names and the following arithmetic illustrate a process; they are not connected-account evidence.
The agency confirms the campaign and period, then distinguishes product recipients from people enquiring on their behalf. The source category does not identify that relationship. It obtains the relevant campaign metrics before treating an account-wide summary as a recommendation for the family-course campaign.
Suppose identified rows contain 300 in cost and 15 recorded conversions, while an undetermined row contains 100 in cost and five conversions. The inclusive acquisition cost is 20. The Unknown row retains those five conversions; excluding them from the narrative would misdescribe the population even if the overall total remained correct.
The analyst therefore reconciles the actual returned key rather than claiming there were no unidentified outcomes. It also investigates a Mixed status using real child context. Neither the word Mixed nor the difference in conversion totals identifies one specific criterion to change.
The owner agrees to retain reach while clarifying downstream lead quality and the inherited setting rationale. A later supported correction would receive its own exact proposal. The example demonstrates a concrete agency outcome without inventing a household diagnosis, mistaking category coverage for complete audience knowledge or turning illustrative numbers into a live edit.
Use creative conversation appropriately
A request for ideas can be answered creatively from the supplied brief. The assistant can suggest a benefit statement, description or offer framing without calling demographic APIs. It should not present those ideas as proven preferences of the Parent category when no creative experiment supports that claim.
Use the client's product context to make the suggestions useful. Availability, service scope and verified benefits can shape copy. Do not invent household needs or family circumstances to personalise a message merely because the reporting discussion concerned parental status.
A request for actual performance requires the appropriate source evidence. Creative brainstorming and metric reporting are different conversational tasks, even when they occur in the same thread. The assistant should choose the necessary work rather than routing every wording question through live account reads.
Drafting copy does not upload an ad. If the client later requests creation or editing, prepare the supported asset operation with its required target and landing-page context. Keep the creative decision and live execution distinct so a casual suggestion does not become an unauthorised write.
When several steps are requested, preserve the real dependencies. The ad needs the correct campaign and ad-group context, while a reporting category does not create that hierarchy. A clear proposal should show the actions the owner can recognise instead of hiding them in a technical payload.
Distinguish audiences, signals and exclusions
The audience builder accepts parental-status ingredients in supported audience preparation. That is not equivalent to editing an existing demographic exclusion. The agency should describe the actual audience effect rather than substituting it for the client's requested eligibility change.
Creating an audience alone leaves it unattached. Creating and associating an audience signal affects the selected asset group. Attaching an existing supported audience retains existing signals, and an already attached audience can produce a no-op. These operations need their own clear target descriptions.
None of those association effects proves that a campaign's existing parental-status criteria were changed. A Performance Max signal also does not expand the current ad-group demographic report into complete Performance Max parental-status performance. Settings capability and reporting coverage remain separate facts.
Establish ingredient validity and resource applicability during preparation. An array in a builder is not proof that every demographic enum or campaign context was already fully validated locally. Provider checks and real operation support still matter before the assistant offers execution.
The owner-facing message should name whether an audience is created, attached or retained and where. Raw dimension JSON is an internal detail. A short, readable explanation can accurately convey the effect without implying that everyone outside the specified parental-status category is excluded.
Prepare a precise supported proposal
The current convenience catalogue does not provide a universal parental-status exclusion fix. A generic schema-mediated action must establish the actual supported resource and field before it becomes executable. Reading a category's performance is not evidence that every imaginable edit is available in the same surface.
Discovery should identify the exact existing target and its real parent context. The confirmation names the campaign, group and intended action where applicable. Asset names keep the message understandable; raw IDs and an invented parent relationship do not help the owner evaluate the proposal.
Present separate logical actions as visible steps in the supported confirmation. If the complete request cannot be prepared within the review safeguards, ask the client to divide it before writing. Do not silently apply only the first steps and describe the whole request as completed. The owner should see every accepted effect and its target.
No live change occurs merely because the assistant discussed a recommendation. The connected owner must confirm the prepared proposal before execution. The visible message should state that no changes have been made yet and that the proposal expires in fifteen minutes.
If preparation cannot establish the requested effect, explain the specific limitation. Retain the useful reporting answer rather than offering an unrelated operation as a substitute. This makes the conversation honest and actionable without abandoning the client's actual objective.
Verify the actual approved effect
After confirmation, show a clear change-in-progress state so the owner knows the click was received. Preflight checks current settings and eligibility before dispatch. That stage protects the approved scope from stale discovery; it does not replace later readback of the result.
Verification compares the resulting configuration with the effect requested. A provider acceptance alone does not establish complete verification, and a successful Slack response alone does not establish a setting change. Report those states accurately rather than collapsing them into one generic success message.
If dispatch may have succeeded but readback remains uncertain, reconcile the focused result. Do not blindly repeat the write because the message was delayed or the verification query failed. An ambiguous mutation needs evidence about what happened before another potentially duplicative operation.
A final executed-and-verified message should be prominent and list the actual completed steps. It should not expose undefined parents, resource identifiers or claim that unrequested settings changed. Readability and technical accuracy matter together when the owner relies on the result.
Technical verification does not guarantee improved commercial performance. The agency still needs a later review using the appropriate population and measurement window. Keep that follow-up separate from the execution status so the client can distinguish a confirmed setting from an outcome still to be measured.
Deliver a useful client report and handoff
Lead with the requested conclusion and the evidence supporting it. State the selected scope, period and metric, then explain any material Unknown reconciliation or availability issue. A compact report can be clear without hiding the conditions that determine whether its recommendation is justified.
Keep displayed metrics to no more than two decimals and use verified currency and ratio labels. Calculate from the retained underlying values. An export should preserve the same population and definitions as Slack instead of turning fallback rows into measured observations or omitting unidentified contributions.
The assistant should retain settled thread context for follow-up questions. Reuse compatible evidence when it already answers the request; obtain fresh settings or another scoped read when necessary. This reduces avoidable API calls without pretending every prior report is current or fits every later comparison.
Finish the handoff with the real decision, remaining evidence and next owner. Record retained reach, commercial questions or the exact verified change as appropriate. Another analyst should understand which campaign and categories were reviewed without reconstructing the whole Slack conversation.
A useful parental-status audit explains advertising evidence rather than guessing family identity. It makes coverage, current settings and commercial intent visible, then chooses the supported next step. That gives the client a practical outcome while keeping the agency's claims within what the actual tools and sources establish.