# DV360 Reach & Frequency Guide ## Enterprise Operating System Guide **Purpose:** Enterprise framework for planning, activating, measuring, optimizing, testing, reporting, and governing reach and frequency across Display & Video 360. **Companion guides:** Use alongside the related operating and specialist guides in this knowledge library. **Core rule:** Impressions are delivery events. Reach estimates distinct people or households exposed under a defined methodology. Frequency describes repeated exposure. Incremental reach describes additional reach beyond a stated comparison. These measures must not be confused. --- # 1. Enterprise Mandate The DV360 Agent must separate seven questions: 1. How many impressions were delivered? 2. How many distinct people, devices, accounts, or households were reached? 3. How often were reached units exposed? 4. How was exposure distributed across the reached population? 5. How much reach was duplicated across tactics or channels? 6. How much additional reach was created beyond a comparison? 7. Did the reach and frequency strategy improve an approved business or brand outcome? Every output must identify the reach unit, method, period, scope, and limitations. --- # 2. Core Principles - Start with the campaign's communications and business objective. - Define the reach unit before using a reach number. - State whether reach is observed, deduplicated, modelled, or estimated. - Analyse frequency distribution, not only average frequency. - Distinguish total reach from incremental or added reach. - Do not sum reach figures that contain duplication. - Do not invent universal effective-frequency ranges. - Use approved benchmark records only. - Evaluate the next investment using marginal reach, not historical average reach. - Separate audience eligibility from delivered reach. - Report quality and business outcomes alongside reach. - Use experiments when the decision depends on causal impact. --- # 3. Reach Decision Tree ## If the objective is broad awareness Prioritise relevant unique reach, controlled exposure, quality, and cost per reached unit. Use brand or search studies where impact evidence is required. ## If the objective is consideration Balance reach with sufficient repeated exposure for the message and decision cycle. Monitor qualified engagement and brand outcomes. ## If the objective is conversion Use reach and frequency as diagnostic controls. Prioritise qualified customers, revenue, or leads rather than maximizing reach alone. ## If the objective is incremental scale Define the comparison campaign or channel, the reach unit, deduplication method, period, and additional-reach metric. ## If reach is low Check audience size, geography, language, device, inventory, creative compatibility, bids, floors, frequency caps, exclusions, and measurement availability. ## If reach is high but outcomes are weak Diagnose audience quality, supply, creative, exposure distribution, destination, measurement, and campaign role before adding spend. --- # 4. Frequency Decision Tree ## Is the reached population concentrated at one exposure? Determine whether this is intended broad coverage or insufficient repetition for the message. Do not assume repeated exposure is required without evidence. ## Is average frequency increasing while reach is flat? Investigate saturation, overlapping line items, limited audiences, narrow supply, deal concentration, device restrictions, and budget growth. ## Are heavy-exposure groups growing? Review frequency distribution, retargeting pools, separate caps, audience overlap, creative fatigue, and marginal outcomes. ## Are caps restricting delivery? Assess reach opportunity, objective, audience size, inventory, bids, formats, and campaign hierarchy before relaxing a cap. ## Do outcomes improve with more observed frequency? Do not infer causality from naturally occurring frequency buckets. Users receiving more impressions may differ from users receiving fewer. Use a controlled frequency experiment where the decision is material. --- # 5. Reach Definitions ## Impressions Total served ad events, including repeated exposures. ## Unique reach Estimated number of distinct users or another defined unit exposed at least once. ## Household reach Estimated number of households exposed under the provider's identity and modelling method. ## On-target reach Reach within an approved audience or demographic definition where measurement supports it. ## Effective reach Reach at or above a stated exposure condition. The exposure condition must be approved for the objective and must not be presented as universal truth. ## Deduplicated reach Reach adjusted to avoid counting the same unit more than once across the stated devices, formats, networks, or channels. ## Incremental reach Reach that is additional to a named comparison under a documented methodology. ## Added reach A product-specific reach measure. Use the provider's exact definition and reporting rules. --- # 6. Frequency Definitions ## Average frequency Impressions divided by reached units under compatible definitions. ## Frequency distribution The proportion or number of reached units receiving at least or exactly specified exposure levels, depending on the report definition. ## Recency Time since a previous exposure or qualifying customer event. ## Frequency cap A delivery control limiting exposure count within a defined time window at the applicable campaign hierarchy level. ## Target frequency A product or bidding goal intended to build toward a specified exposure level where supported. ## Saturation A condition where additional spend increasingly creates repeated exposure or weaker marginal value rather than useful new reach or outcomes. --- # 7. DV360 Measurement Framework DV360 unique reach and frequency metrics measure people shown an ad and how often, account for cross-device behaviour, and can account for co-viewing on connected television. Reported metrics include Total Reach, Average Impression Frequency per User, a seven-day average-frequency metric, and frequency-distribution metrics. Unique reach uses statistical models and aggregated signals to deduplicate audiences across sessions, formats, networks, and devices. ## Enterprise reporting requirements - Metric exact name - Reach unit - Provider - Modelled or observed treatment - Co-viewing treatment - Date range - Campaign and inventory scope - Market availability - Reporting limitations Do not describe modelled reach as a deterministic user-level count. --- # 8. Sources of Truth ## DV360 Use for supported campaign reach and frequency, frequency distribution, campaign controls, and product-specific added-reach measures. ## Campaign Manager 360 Use for supported ad-server exposure and cross-channel reporting according to the implemented identity and methodology. ## YouTube reporting Use for supported YouTube reach, frequency, format, and campaign measures using current product definitions. ## Third-party cross-media measurement Use when the selected provider supports the required digital, television, or cross-platform comparison. State provider methodology and market coverage. ## CRM, analytics, and business systems Use to evaluate whether reached users or campaigns contribute to qualified engagement, customers, revenue, or other outcomes. These systems may not provide media-deduplicated reach. The source depends on the question. --- # 9. Reach Measurement QA Before using a reach figure, confirm: - Exact metric - Unit of reach - Date range - Deduplication scope - Included formats and inventory - Device coverage - Co-viewing treatment - Modelled component - Market availability - Reporting latency - Minimum availability conditions where applicable - Whether daily figures can be added Reach availability can depend on country and sufficient impression and unique-user data. Most reach and frequency metrics have a limited reporting date range. Verify current limits in official documentation rather than hardcoding them into planning templates. --- # 10. Reach Curve Framework A reach curve shows how estimated unique reach changes as spend or impressions increase. ## Use cases - Compare investment scenarios - Identify slowing marginal reach - Evaluate audience or inventory expansion - Support frequency decisions - Inform budget reallocation ## Interpretation - A steep section indicates additional investment is reaching many new units. - A flattening section indicates more investment is creating proportionally less new reach. - A curve can shift when audience, inventory, format, device, market, identity, or creative eligibility changes. ## Limitations A reach curve does not prove quality or business impact. --- # 11. Marginal Reach Framework Average cost per reached unit describes the full historical campaign. Marginal cost per additional reached unit describes the efficiency of the next investment. ## Evidence inputs - Spend by period or controlled increment - Unique reach - Added or incremental reach where supported - Impressions - Frequency distribution - Audience size - Inventory availability - Bids and floors - Quality - Qualified outcomes ## Saturation signals - Spend rises faster than reach. - Frequency rises while reach is stable. - Heavy-exposure groups grow. - New inventory is lower quality. - Cost per additional reached unit increases. Do not claim a precise marginal return when available data cannot isolate the next increment. --- # 12. Reach Efficiency Framework Potential calculations include: - Cost per reached user or household - Cost per added reached unit - Cost per on-target reached unit - Cost per qualified reached unit where a valid method exists Every calculation must align cost and reach scope, period, currency, and methodology. Qualify comparisons across markets, formats, providers, and reach units. --- # 13. Frequency Distribution Framework Average frequency can hide unequal exposure. Analyse the distribution using the supported report. Possible questions: - How much of the reached population received only one exposure? - How much received repeated exposure? - Where are heavy-exposure groups concentrated? - Which audiences, markets, devices, publishers, deals, or line items drive repetition? - Does repeated exposure correlate with business outcomes? - Is the relationship causal or observational? Use current supported fields and exact definitions for frequency-distribution metrics covering multiple exposure levels. --- # 14. The Average-Frequency Problem Two campaigns can have the same average frequency and very different distributions. Example interpretation only: - Campaign A may expose most reached units several times. - Campaign B may expose many once and a small group very heavily. The average alone cannot distinguish these patterns. Reporting should include: - Average frequency - Distribution - Reach trend - Heavy-exposure concentration - Audience and supply context - Business outcome Do not invent exposure bands as mandatory enterprise categories. Use supported report levels and approved analytical groupings. --- # 15. Effective Frequency Governance Effective frequency is a planning or analysis construct linking reach to an exposure condition. The condition should reflect: - Objective - Message complexity - Creative strength - Format - Decision cycle - Audience familiarity - Campaign length - Competitive environment - Evidence from tests or approved benchmarks A universal "ideal frequency" is not supported across all campaigns. When evidence is limited, use scenarios and controlled testing rather than a hard rule. --- # 16. Frequency Cap Architecture Frequency controls can exist at different levels of the campaign hierarchy where supported. The Setup & Trafficking Guide should govern where caps are applied. This guide governs the decision logic. ## Questions - Is the cap intended to control the full campaign, an insertion order, or a tactic? - Are several caps applying simultaneously? - Can a user receive exposure from another campaign outside the cap? - Do YouTube, CTV, display, video, audio, or deals share the intended control? - Does the cap period match the planning and reporting period? - Is the control protecting reach or preventing useful repetition? Verify current product behaviour before implementation. --- # 17. Added Reach from Frequency Capping DV360 can report additional users reached because of a campaign-level frequency cap: total unique users and added unique users. Added reach is not additive because people can be duplicated across days. The reporting period should be at least as long as the frequency-cap period and reach data requires stabilization before final reporting. Verify current reporting windows and latency in the product documentation. ## Interpretation rules - Added reach is relative to the configured cap and product methodology. - It is not automatically reach incremental to another channel. - Do not add daily added-reach values. - State the cap, reporting period, and campaign scope. --- # 19. Audience Reach Framework Separate: - Addressable or theoretical population - Eligible platform population - Matched audience - Reachable population under campaign settings - Delivered unique reach - Qualified or on-target reach where measured Loss can occur through: - Consent and data availability - Identity matching - Market support - Inventory compatibility - Geography and device - Bid and budget - Frequency controls - Exclusions Do not report addressable audience size as delivered reach. --- # 20. Audience Overlap Framework Overlap can create duplicated exposure and internal auction competition. Review: - Shared audience membership - Prospecting and remarketing overlap - Existing-customer overlap - Multiple line items targeting the same population - Cross-format exposure - Cross-campaign exposure - Different identity methods ## Actions - Consolidate redundant audiences. - Apply exclusions where strategically required. - Coordinate caps. - Preserve distinct tests. - Accept overlap when intentional. Audience-list overlap is not the same as measured reach duplication. State the method. --- # 21. Audience Exhaustion Framework Potential signals: - Reach growth slows. - Frequency rises. - Heavy-exposure groups grow. - CPM or cost per additional reach increases. - Creative response weakens. - Remarketing pools stop growing. Potential actions: - Expand permitted audiences. - Add suitable inventory or formats. - Coordinate caps. - Refresh creative. - Reduce spend. - Test new markets or contexts. Do not diagnose exhaustion solely from average frequency. --- # 22. Supply Reach Framework Assess supply by: - Unique reach contribution - Added or incremental reach where valid - Duplication with other supply - Cost - Quality - Audience relevance - Format and device - Publisher or app transparency - Qualified outcomes Open Auction may provide broad scale. Curated supply may provide controlled context or quality. Deals may provide access, priority, or certainty. None is inherently superior for reach without evidence. --- # 23. PMP and Programmatic Guaranteed Reach Evaluate a deal using: - Unique inventory - Reach contribution - Duplication - Frequency concentration - Cost per additional reach - Delivery certainty - Quality - Contractual commitment - Business outcomes A separate reporting view can exist for Programmatic Guaranteed deals. Verify the current product workflow and definitions before use. Measure whether a premium buys unique reach or certainty. --- # 24. Open Auction and Curated Supply Reach ## Open Auction Review scale, long-tail distribution, duplicate supply paths, quality, audience reach, device mix, and marginal reach. ## Curated supply Review whether curation reduces waste, improves quality, or reaches a valuable context while preserving enough scale. An omnichannel reach approach can span display, video, CTV, audio, and digital out-of-home, with frequency and exposure managed across channels and quality-controlled environments. This describes strategic intent, not approved numeric reach benchmarks. --- # 25. YouTube Reach Framework For YouTube activity, document: - Campaign and line-item type - Format - Device - Audience - Inventory setting - Unique reach - Frequency and distribution - Co-viewing treatment - Added reach where supported - Brand or business outcome A target-frequency capability can build toward a weekly target across coordinated YouTube line items through insertion-order settings. As feature scope and availability can change, verify current official documentation before implementation. Do not assume every YouTube format produces the same reach behaviour. --- # 26. Connected Television Reach Framework CTV reach may be expressed through users, devices, households, or modelled co-viewers depending on the method. Review: - App and publisher identity - Device or household unit - Co-viewing - Cross-device deduplication - Frequency - Inventory and deals - Quality - Relationship with linear television - Incremental reach methodology - Creative and QR or follow-on journey DV360 unique-reach metrics can account for co-viewing on connected television devices. Do not treat completed views as unique household reach. --- # 28. Cross-Channel Reach and Deduplication Cross-channel reach requires a method that can identify or model duplication across the stated channels. Document: - Channels included - Unit of reach - Identity or model - Observed and modelled components - Time period - Market coverage - Co-viewing treatment - Unsupported inventory Without a credible deduplication method, report channel reach separately and state that totals cannot be summed into unique reach. Unified frequency and coordinated campaign setup can reduce duplication and expand reach, but this is an operating objective rather than proof that all cross-channel exposure is deduplicated. --- # 29. Incremental Reach Decision Tree 1. Additional to what comparison? 2. What is the reach unit? 3. Which channels, campaigns, or formats are included? 4. Is deduplication available? 5. Is the comparison simultaneous and methodologically consistent? 6. Is the exposed population relevant? 7. Is the added reach high quality? 8. Does additional reach improve brand or business outcomes? 9. What is the cost per additional reached unit? 10. Does the next investment still add reach efficiently? If the comparison or deduplication method is missing, do not use "incremental reach." Use "additional platform-reported reach" only when that exact product definition applies. --- # 30. Reach Experimentation Playbook Test questions can include: - Does another format add unique reach? - Does a PMP reach users not available through Open Auction? - Does a broader audience create useful additional reach? - Does a different cap improve added reach? - Does CTV create reach beyond online video or television? ## Design requirements - Business decision - Comparison - Reach unit - Control and treatment where causal evidence is required - Deduplication method - Primary outcome - Quality and business guardrails - Period and lag - Contamination - Scale decision Use the Testing & Experimentation Guide for full governance. --- # 31. Reach-to-Outcome Framework Reach is an intermediate media outcome. Link it to: - Brand awareness or consideration - Search behaviour - Qualified sessions - Leads and customers - Revenue and margin - Incremental outcomes Questions: - Was the intended audience reached? - Did exposure quality meet policy? - Was frequency appropriate? - Did outcomes improve as reach grew? - Did marginal outcomes weaken as frequency increased? - Was the relationship causal or observational? Do not report reach growth as business success without connecting it to the approved objective. --- # 32. Quality-Adjusted Reach Governance A quality-adjusted or attention-adjusted reach concept requires an approved methodology. Document: - Reach source - Quality or attention provider - Eligible overlap population - Formula - Weighting - Unsupported inventory - Benchmark status - Owner and approver - Relationship with business outcomes Quality-controlled exposure and attention-adjusted reach are desired success signals, but without a complete approved formula, this guide does not invent one. --- # 34. Optimization Playbook ## Reach below plan and quality healthy Check budget, pacing, bids, audience, inventory, creative compatibility, frequency caps, deals, and exclusions. ## Reach below plan and quality weak Correct quality, audience, creative, supply, or measurement before expanding. ## Frequency above plan Check audience size, overlap, caps, budget, retargeting, inventory concentration, and creative rotation. ## Marginal reach slowing Consider budget reduction, audience expansion, new formats, additional supply, coordinated caps, or a controlled test. ## Reach healthy and business outcomes weak Diagnose relevance, creative, destination, customer quality, attribution, and incrementality. More reach may not solve the issue. --- # 35. Root-Cause Scenario Library ## Reach falls despite higher budget Check rising CPM, narrower audiences, inventory loss, creative incompatibility, higher frequency, deals, device shift, and measurement changes. ## Frequency rises without reach growth Check saturation, small retargeting pools, duplicated audiences, limited supply, separate caps, and budget increase. ## CPM rises while reach slows Check auction pressure, inventory restriction, quality controls, deals, audience precision, device, and diminishing marginal access. ## Reach improves while CPA worsens Check audience quality, inventory expansion, creative relevance, landing journey, attribution, and conversion lag. ## Reach improves while quality declines Identify newly accessed sites, apps, devices, exchanges, publishers, and audience expansion. ## CTV adds reach but conversions remain weak Confirm the campaign role, household or user method, attribution, creative, frequency, search or brand outcomes, and incrementality. ## YouTube added reach declines Check cap changes, campaign scope, reporting period, audience overlap, saturation, format mix, and product reporting conditions. ## Reach is duplicated across channels Check whether a cross-channel identity method exists, whether standalone campaigns are outside shared controls, and whether audiences overlap. ## Audience appears exhausted Review reach curve, frequency distribution, audience growth, creative fatigue, CPM, and marginal outcomes. ## Frequency cap is too restrictive Check underdelivery, low repeated exposure, campaign objective, audience size, inventory, and test evidence before relaxing. ## Frequency cap is too loose Check heavy-exposure groups, reach loss, customer experience, creative fatigue, and marginal outcomes. ## Reach curve flattens early Check narrow targeting, limited inventory, high competition, creative sizes, geography, device, deals, and measurement availability. ## Reach grows while sales remain flat Check whether reach is relevant, whether the campaign role is awareness, customer decision length, creative, offer, distribution, and causal evidence. ## Reach declines after PMP activation Check whether the budget shifted into duplicated or narrower inventory, deal availability, price, audience, and frequency. ## Open Auction repeats the same users Check audience size, supply concentration, device, caps, bid strategy, and whether additional inventory adds unique reach. ## Curated supply reduces reach Assess whether quality, context, certainty, or outcomes justify the reduced scale and higher concentration. ## Incremental reach is high but conversions are low Check audience relevance, message, destination, campaign role, conversion lag, and whether conversion was the intended outcome. ## Incremental reach is low but conversion quality is high The tactic may be valuable as demand capture rather than reach extension. Judge it by its approved role. ## Reach grows while viewability falls Check added inventory, apps, placements, formats, devices, and measured coverage. ## Reach grows while suitability declines Review whether audience or supply expansion accessed different contexts and whether pre-bid coverage remains effective. ## Average frequency is stable but complaints rise Inspect distribution, heavy-exposure pockets, market or audience concentration, creative repetition, and cross-campaign exposure. ## Reach differs between DV360 and third party Align reach unit, period, identity, devices, channels, co-viewing, modelling, and unsupported inventory. Do not average the values. ## Daily reach totals exceed period reach The same people can appear on multiple days. Daily unique reach is not necessarily additive. ## Added reach appears lower over a longer report Duplicate people across days mean added reach is not additive and can decline over longer periods. Validate the report range and product definition. --- # 36. Reach QA Template ```markdown # DV360 Reach QA - Report or plan ID: - Business question: - Campaign and market scope: - Reach source: - Exact metric: - Reach unit: - Date range: - Inventory and format scope: - Device scope: - Deduplication method: - Co-viewing treatment: - Modelled component: - Added or incremental comparison: - Measured and unsupported population: - Currency and cost basis: - Reporting latency considered: - Daily additivity checked: - Benchmarks and status: - Reviewer: - Approver: - Publication decision: ``` --- # 37. Frequency QA Template ```markdown # DV360 Frequency QA - Campaign and period: - Frequency source: - Average-frequency metric: - Distribution metrics: - Reach unit: - Cap hierarchy: - Cap count and time window: - Target-frequency setting where used: - Cross-campaign exposure limitation: - Audience and market concentration: - Heavy-exposure finding: - Creative rotation: - Business and quality guardrails: - Review trigger: - Responsible owner: - Accountable approver: ``` --- # 38. Reach & Frequency RACI ```markdown # Reach and Frequency RACI | Activity | Responsible | Accountable | Consulted | Informed | |---|---|---|---|---| | Define reach objective | | | | | | Select reach source and unit | | | | | | Approve planning assumptions | | | | | | Design campaign architecture | | | | | | Configure frequency controls | | | | | | Validate reach reporting | | | | | | Analyse frequency distribution | | | | | | Measure duplication | | | | | | Evaluate added or incremental reach | | | | | | Approve cap change | | | | | | Approve budget reallocation | | | | | | Complete QA | | | | | | Approve report | | | | | | Update methodology registry | | | | | ``` Assign one accountable owner for methodology and one for each campaign decision. --- # 39. Change Control Framework Material changes include: - Reach source or methodology - Reach unit - Deduplication scope - Co-viewing treatment - Frequency cap or target - Campaign hierarchy affecting shared exposure - Audience or inventory expansion - Added-reach comparison - Benchmark or planning assumption Every change must record: - Current and proposed state - Reason and evidence - Affected campaigns and reports - Expected reach and frequency effect - Quality and business trade-off - Historical comparability - Owner and approver - QA, review trigger, and rollback --- # 40. Change Control Template ```markdown # Reach and Frequency Change Request - Change ID: - Campaign scope: - Current state: - Proposed state: - Reach or frequency issue: - Evidence: - Methodology impact: - Expected reach effect: - Expected frequency effect: - Cost and quality trade-off: - Business impact: - Historical comparability impact: - Responsible owner: - Accountable approver: - QA reviewer: - Implementation window: - Monitoring metrics: - Review trigger: - Rollback condition: - Resolution evidence: ``` --- # 41. Methodology Registry ```markdown # Reach and Frequency Methodology Registry | Method ID | Status | Market | Channel Scope | Reach Unit | Deduplication | Co-Viewing | Frequency Method | Effective Date | Review Date | Owner | Approver | |---|---|---|---|---|---|---|---|---|---|---|---| | No approved records | Unavailable | N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A | ``` ## Full Record ```markdown ## Reach and Frequency Method: [ID] - Status: - Business question: - Provider and source: - Market: - Channels, formats, and inventory: - Reach unit: - Deduplication method: - Identity or modelling method: - Co-viewing treatment: - Average-frequency definition: - Distribution metrics: - Added or incremental reach definition: - Reporting period rules: - Latency and availability: - Permitted use: - Limitations: - Effective date: - Review date: - Owner: - Approver: - Previous version: ``` --- # 42. Audit Framework Audit: - Objective and business question - Reach source and unit - Deduplication - Co-viewing - Date and campaign scope - Frequency hierarchy - Frequency distribution - Audience overlap - Supply contribution - Added or incremental reach - Quality and business outcome - Benchmarks and assumptions - Change history - QA and ownership Classify: - Retain - Correct - Consolidate - Expand - Test - Reclassify - Retire - Escalate --- # 43. Maturity Model ## Level 1: Impression buying Reach and repetition are inferred mainly from impressions. ## Level 2: Reach reporting Unique reach and average frequency are reported with basic definitions. ## Level 3: Governed exposure Distribution, deduplication, QA, RACI, methodology registry, and change control are established. ## Level 4: Incremental reach optimization Cross-format and cross-channel evidence guides caps, audiences, supply, and investment. ## Level 5: Marginal reach planning Reach curves, causal testing, quality, business outcomes, and marginal contribution guide portfolio allocation. Recommend the next practical improvement rather than complexity for its own sake. --- # 44. Prohibited Behaviours Never: - Treat impressions as unique reach. - Treat cookies or devices as people without stating the method. - Treat average frequency as sufficient. - Sum daily or channel reach without valid deduplication. - Claim incremental reach without naming the comparison and method. - Invent identity, co-viewing, or deduplication logic. - Apply a universal effective-frequency threshold. - Relax mandatory quality controls solely to expand reach. - Claim observed frequency caused an outcome without a valid design. - Treat completed views as household reach. - Hide modelled or unsupported populations. - Select the largest reach source as the truth. - Scale budget from average reach efficiency without assessing marginal reach. - Claim a live setting or methodology was implemented without evidence. --- # 45. Default Agent Output **Mode selected:** Reach Plan / Frequency Strategy / Campaign Review / Incremental Reach / Optimization / Audit ## Business question ## Source and methodology ## Reach analysis ## Frequency analysis ## Distribution and saturation ## Duplication analysis ## Added or incremental reach ## Quality and business outcomes ## Findings ## Recommendation ## Expected trade-off ## QA status ## RACI ## Change-control status ## Confidence ## Review trigger and rollback --- # 46. Final Enterprise Instruction Reach and frequency are governed exposure decisions, not secondary reporting metrics. The DV360 Agent must define the reach unit and methodology, analyse the full frequency distribution, identify duplication and saturation, distinguish added reach from cross-channel incremental reach, connect exposure with quality and business outcomes, and evaluate the next investment through marginal contribution. When deduplication, co-viewing, or incrementality cannot be established, the Agent must disclose the limitation rather than invent a precise reach claim.