How MediaProof Labs Evaluates Your Media Plan
MediaProof Labs combines a structured media-review framework with AI-supported analysis to provide an independent assessment of a media plan. The framework is informed by more than 20 years of digital marketing experience.
The goal isn't to find something wrong with every plan. The goal is to determine whether the plan makes sense for what you're actually trying to accomplish.
Context comes first
There is no universally correct media mix. A heavy paid search investment may make perfect sense for one business and create strategic risk for another. We consider the context you provide:
- Objective
- Business model
- Audience
- Geography
- Budget
- Campaign duration
- KPI
- Funnel
- Historical information you choose to provide
We don't assume a plan is good simply because it follows common industry practices. And we don't assume an unconventional plan is bad simply because it looks different.
The areas we review
Strategy & Objective Alignment
Does the media strategy support the stated business objective?
Audience Strategy
Is the plan designed to reach the right people? We consider target audience definition, addressable audience, geography, audiences, prospecting and retargeting, overlap, exclusions, ABM where relevant, and whether targeting may be too narrow or too broad.
Reach & Frequency
When sufficient information is available, we evaluate unique reach, percentage of audience reached, average frequency, duration, frequency by channel, cross-channel frequency, duplication, incremental reach, caps, and potential under- or over-exposure. There is no universal correct frequency.
Share of Voice
When relevant and supported, we consider estimated share of voice, competitive activity, category saturation, market position, growth objectives, and geographic concentration. We do not invent competitor spending or market data when it isn't available.
Budget Allocation & Media Economics
We evaluate where the budget is going and how much actually reaches media — distinguishing working media from agency fees, planning fees, platform/DSP/ad-serving/data/measurement fees, creative production, technology, commissions, markups, and other costs. Non-working media is not automatically waste.
Fees, Commissions & Markups
We look for costs that may not be obvious from the top-line budget — agency commissions, percentage-of-spend management fees, technology/DSP/data fees, ad-serving, programmatic supply-chain costs, production markups, and pass-through costs. We do not assume a fee is inappropriate simply because it exists.
Media Quality & Viewability
Where relevant information is available, we consider served vs. measurable vs. viewable impressions, viewability rate, invalid traffic, brand safety and suitability, placement quality, inventory type, and video/CTV completion. We distinguish served from viewable impressions when the information allows.
Channel Mix & Funnel Coverage
We evaluate whether channels have clear roles across demand creation, demand capture, prospecting, retargeting, nurture, conversion, and retention or expansion. More channels do not automatically create a better plan.
Forecast Integrity
We pressure-test the assumptions connecting investment to expected results — Spend→Impressions→Clicks→Conversions→Leads→Opportunities→Pipeline→Revenue — and identify mathematical inconsistencies and unsupported assumptions.
Testing & Learning
We evaluate whether the plan creates meaningful opportunities to learn through creative, audience, channel, offer, and landing-page testing and structured experimentation. We don't recommend testing simply for the sake of testing.
Measurement & Attribution
We evaluate whether the organization will actually be able to determine whether the investment worked — KPI definitions, conversion tracking, CRM integration, offline conversions, pipeline/revenue measurement, attribution, incrementality, brand and conversion lift, MMM where appropriate, platform-reported vs. independently measured results, and reporting cadence.
Not every finding is the same
Clearly classify findings as:
We don't invent benchmarks
We do not use generic benchmarks simply because they are available. When external evidence informs a finding, we identify the source when possible. Where a reliable benchmark doesn't exist for the specific situation, we say so.
Confidence matters
High Confidence
Strongly supported by available information.
Medium Confidence
Evidence supports the conclusion, but additional context could materially affect it.
Low Confidence
Directional assessment requiring additional information.
AI supports the analysis. It doesn't make it infallible.
AI helps apply the framework, perform calculations, identify patterns, and evaluate information consistently. It can also be wrong. MediaProof Labs is intended to provide another informed perspective, not replace professional judgment. The detailed scoring logic, decision rules, prompts, weighting, and underlying analysis framework are proprietary and are not published.
