Campaign Impressions Analysis

Goal of the analysis:

Quantify how effectively a campaign delivered impressions across channels, audiences, and time, and determine the quality and usefulness of those impressions (reach, frequency, viewability, on-target delivery, and invalid traffic). For executives, this analysis ensures media dollars are translating into real, in-view exposures among priority audiences, highlights waste (over-frequency, low viewability, off-target), and informs optimizations to pacing, channel mix, and creative deployment to maximize efficient reach and downstream impact.

Data required:

• Media plan and goals:
  • Planned impressions by channel/placement, budget, CPMs, flight dates, daily/weekly pacing targets.
  • Target audience definitions (demographics, interests, geo), frequency caps, brand safety requirements.
  • KPIs: reach, average frequency, effective frequency (1+, 3+, 5+), viewability thresholds, on-target %.
• Ad delivery and cost data:
  • Ad server logs (e.g., CM360, Flashtalking, Sizmek): served impressions, clicks, creative IDs, placement IDs, timestamps.
  • Platform reports (DV360, The Trade Desk, Google Ads/YouTube, Meta, TikTok, Snap, X, LinkedIn, Amazon Ads, CTV/OTT platforms): impressions, spend, CPM, reach (where available), frequency, geo/device breakdowns.
  • Traditional media equivalents (TV/Radio/OOH): GRPs/TRPs, estimated impressions, reach/frequency, daypart, DMA.
• Quality and verification data:
  • Viewability and invalid traffic (IVT) from IAS, DoubleVerify, Moat: measurable impressions, viewable impressions, viewability rate, IVT (GIVT/SIVT), brand safety incidents.
  • On-target reach/OTR (Nielsen DAR, Comscore vCE) and deduplicated cross-platform reach where available.
• Audience and context:
  • Geo (country, DMA/region), device (desktop, mobile app/web, CTV), environment (in-app, web, UGC, premium).
  • Inventory details: domains/apps lists, exchanges, PMPs/PG deals, ad sizes and video lengths.
• Historical and benchmark data:
  • Prior campaigns’ impression delivery, viewability, IVT, on-target %, reach/frequency distributions, and vCPM.
  • Platform/category norms for viewability, IVT, and on-target delivery.

Detailed step-by-step instruction on how to conduct the analysis:

1. Align on objectives and definitions. Confirm impression-related KPIs: total and daily/weekly delivery, reach, average and effective frequency (1+, 3+, 5+), on-target rate (OTR), viewability (per MRC: display 50% pixels 1s; video 50% 2s), IVT threshold, and vCPM targets.
2. Extract delivery data. Pull ad server and platform-level reports at the most granular level permitted (campaign/ad group/creative/placement by day). Include impressions, reach (if available), cost, geo/device, and inventory details. For TV/Radio/OOH, obtain GRPs, impressions, and daypart.
3. Ingest verification data. From IAS/DV/Moat, obtain measurable impressions, viewable impressions, viewability rate, and IVT counts by placement/day. From DAR/vCE, pull on-target impressions and deduplicated reach where available.
4. Normalize and map taxonomy. Standardize campaign names, IDs, channel labels, markets, time zones, and creative IDs across sources. Map placements to a consistent channel taxonomy (Search, Social, Display, Online Video, CTV, DOOH, TV, Audio).
5. Quality filter and compute core metrics.
   • Total impressions = sum of served impressions.
   • Measurable impressions and viewable impressions per MRC; viewability rate = viewable/measurable.
   • IVT rate = invalid impressions/served impressions; Valid impressions = served − IVT.
   • vCPM = spend / viewable impressions × 1000; Cost per on-target impression = spend / on-target impressions.
   • Delivery variance % = (Actual − Planned) / Planned; Pacing index = cumulative actual / cumulative plan.
6. Reach and frequency analysis. Where available, compute reach and frequency per channel/overall. Average frequency = impressions / reach. Create frequency distributions (1, 2–3, 4–5, 6–9, 10+). Calculate effective reach 1+, 3+, 5+ (share of audience receiving at least N exposures). For TV/radio, convert GRPs to impressions and use GRP = Reach% × Avg Frequency to sanity-check.
7. Segmentation. Break metrics by market/DMA, device (desktop/mobile/CTV), inventory type (open exchange vs PMP/premium), daypart, creative length/size, audience segment, and placement. Identify pockets of under-delivery, over-frequency, low viewability, or high IVT.
8. Pacing diagnostics and forecasting. Plot daily/weekly impressions vs. plan. Identify early/late pacing, flighting gaps, and learning-phase lags. Forecast end-of-flight delivery using run-rate or simple time-series; recommend budget reallocation to hit goals.
9. Overlap and deduplication. Where cross-platform dedup reach is available (DAR/vCE/platform modeled reach), estimate duplication and compute deduped reach and average frequency. If not available, triangulate with platform reach and apply conservative overlap assumptions from historical data.
10. Inventory and brand safety review. Evaluate domain/app lists, supply paths (SSPs), PMPs, and placement categories. Flag low-quality supply (low viewability, high IVT, brand safety issues) and identify candidates for exclusion/whitelisting or SPO (supply-path optimization).
11. Benchmarking and historical comparison. Compare current campaign to internal quartiles and external norms for viewability, IVT, OTR, vCPM, and effective frequency. Highlight material deviations and their budget impact.
12. Insight synthesis and actions. Quantify waste (impressions lost to IVT, non-viewable, off-target, over-frequency). Prioritize actions by impact: frequency caps, reallocation to high-OTR/high-viewability inventory, creative/format shifts, geo/device optimization, and deal renegotiations.
13. Governance and documentation. Store methods, exclusion lists, PMP contracts, and benchmark tables. Set a cadence for mid-flight checks (e.g., 2–3x weekly on large flights).

Format of the output of analysis:

• Executive summary slide: plan vs. actual impressions, reach, frequency, viewability, IVT, OTR, vCPM; top three issues and actions.
• Pacing dashboard: daily/weekly delivery vs. plan with forecast to end-of-flight.
• Quality waterfall: served → valid (−IVT) → measurable → viewable → on-target impressions with loss percentages and cost impact.
• Reach/frequency report: average frequency and frequency distribution; effective reach 1+/3+/5+ by channel.
• Segment heatmaps: viewability, IVT, vCPM by channel, inventory type, geo, and device.
• Inventory table: domains/apps with performance, brand safety flags, and recommended include/exclude lists.
• Benchmark comparison: internal quartiles and external norms with variance indicators.

How to interpret results:

• High impressions with low reach (high frequency): Indicates over-concentration; expect diminishing returns and potential annoyance. Tighten frequency caps or broaden audience/inventory.
• Low viewability or high IVT: A large share of spend is not producing real, in-view exposures. Shift to higher-quality supply (PMPs/premium), enable viewable bidding, and enforce exclusions.
• Low on-target rate: Targeting precision or third-party audience quality is weak; refine segments, data partners, or geo/device filters.
• vCPM vs CPM: A low CPM with poor viewability can produce a high vCPM (inefficient). Optimize toward vCPM or cost per on-target impression.
• Pacing deviations: Under-delivery risks missing reach goals; over-delivery early can inflate frequency. Adjust bids, budgets, and learning phases promptly.
• Cross-channel differences: Social/UGC may deliver scale with variable viewability; CTV typically has high viewability but higher CPM—assess cost per effective reach.
• Trends and volatility: Stable or improving quality metrics indicate healthy supply. Spikes in IVT/viewability drops often coincide with new inventory; audit immediately.

Steps a company can take to improve on this measure:

• Pacing and delivery management:
  • Set daily pacing and alert thresholds (e.g., ±10% vs plan). Use automated rules to shift budget between line items.
  • Stagger creatives and expand inventory earlier to avoid end-of-flight compression and over-frequency.
• Quality and supply optimization:
  • Adopt viewable CPM or outcome-optimized bidding; enforce MRC+ viewability thresholds (e.g., 70%+ for display, 80%+ for video).
  • Implement supply-path optimization; prioritize PMPs/private deals and high-quality publishers; exclude low-viewability/high-IVT domains/apps.
  • Apply third-party verification pre-bid segments (brand safety, fraud) and maintain active block/allow lists.
• Targeting and reach expansion:
  • Broaden audience pools and incrementally layer lookalikes/contextual to grow unique reach.
  • Tune geo/device mix (e.g., increase CTV for quality exposures; rebalance from in-app gaming if IVT is elevated).
• Frequency and creative strategy:
  • Set platform-specific frequency caps and cross-platform guardrails; monitor heavy-frequency tails (10+ exposures).
  • Rotate creatives and vary formats to mitigate wearout; align video lengths to placement norms to improve completion/viewability.
• Data, systems, and governance:
  • Implement consistent ID/taxonomy across platforms; integrate verification data into dashboards.
  • Use cross-channel reach tools (e.g., DAR, vCE, platform-modeled reach) and establish internal effective-reach benchmarks.
  • Run mid-flight quality audits and quarterly inventory reviews with agencies and SSPs.
• Scenario guidance:
  • If impressions are on plan but reach is low, increase audience breadth, add new inventory, and tighten frequency caps.
  • If vCPM is high due to low viewability, switch to viewable bidding, migrate to PMPs, and refine creative/load times.
  • If OTR is low, reassess audience data partners and refine age/geo/device targeting; consider contextual where IDs are limited.

Benchmark comparisons:

General benchmarks:

• Viewability (MRC): Display 50–70% typical; Video 65–85%; CTV often 90%+ (measurable permitting). Aim to exceed internal top quartile.
• Invalid Traffic (IVT): Keep below 1–3% of served impressions; spikes merit immediate supply review.
• On-Target Rate (OTR): 35–60% is common depending on data quality and platform; premium data/placements can push higher.
• Effective frequency: For awareness, 1–3+ exposures is a common starting target; for consideration, 3–5+. Validate with brand lift where possible.
• Pacing: Maintain daily delivery within ±10% of plan for steady flights; larger tolerance during learning phases.

Segment- or industry-specific benchmarks:

• Premium publishers/PMPs: Higher viewability (70–90%) and lower IVT than open exchange; CPMs higher but vCPM often comparable or better.
• UGC/social environments: Large scale; viewability varies by format (stories/reels vs feed). Expect broader reach with careful frequency controls.
• CTV/OTT: High viewability and on-target potential; CPMs higher; monitor ad pod frequency and duplication across apps.
• TV/Radio/OOH: Use GRP/TRP norms for the category; translate to impressions and effective reach targets using historical reach curves.
• When external norms are sparse, establish internal benchmarks: compare to prior similar flights, top-quartile placements, and best-performing markets; set quarterly improvement goals for vCPM, OTR, and viewability.