Before You Trust an Earned Media Value Number, Ask These Seven Questions
A buyer-side test for deciding whether an EMV report is decision-ready, merely directional or too opaque to use.
Analysis by Tu Dang · Founder of ROLearn
Published · 7-minute read

Key findings
- An earned media value estimate is only as credible as its definitions, evidence boundaries, paid-versus-earned policy and market comparables.
- A result that mixes views, followers, visits and readership before normalization cannot be audited from the total alone.
- The absence of an audience-overlap policy and uncertainty range is a material reporting gap, not a technical footnote.
An earned media value number often arrives at the end of a campaign deck in large type. That presentation encourages the wrong first question: is the number high enough?
The better first question is whether the number can survive seven ordinary questions from a CMO, procurement lead or finance partner. If the reporting team cannot answer them, the estimate may still be a directional signal. It is not yet a decision-grade claim.
This field note is a short companion to ROLearn’s full earned media value methodology and its downloadable Earned Media Value Evidence Standard. It does not reveal OMNI-EMV’s proprietary coefficients. It explains the public evidence questions every model should be able to answer.
1. What decision is the number meant to support?
An EMV estimate can help compare earned-attention patterns, diagnose which channels amplified an activation or set a planning benchmark. It cannot answer every question about brand impact, revenue and ROI.
Ask the reporting team to finish one sentence: we will use this estimate to decide… If the answer moves from media equivalence to sales causality, the team needs a separate outcome design and counterfactual.
2. What exactly counted as earned?
Creator partnerships, boosted posts, owned social publishing and organic fan content can appear around the same activation. They are not the same evidence.
A defensible report states how paid, owned and earned media were classified and keeps purchased delivery visible. Otherwise a brand can pay for creator reach once through the contract and count it again as earned value.
3. Which media objects entered the ledger?
The evidence should resolve to inspectable objects: a game event, post, video, stream, article or syndicated item. Each needs an identifier, timestamp, channel and attribution reason.
This matters because monitoring systems often encounter the same content more than once. An article may be syndicated, a clip embedded and a post reposted. Without object-level identity, deduplication becomes guesswork.
4. Were channels valued in the way people consumed them?
A Roblox visit, a video view, a creator livestream minute and a press article readership estimate are different signals. Adding their raw counters produces a total with no stable meaning.
The model should keep each channel in its native unit long enough to qualify the evidence. Only then should it normalize comparable attention to a common reporting basis.
5. How were paid delivery and audience overlap handled?
The same person can enter a branded world, watch a creator and see a social clip. That is valuable frequency, not three unique people.
Ask whether overlap was observed, modeled or bounded. If no identity or cross-media evidence exists, the report should use a conservative assumption and show how the result changes under plausible overlap scenarios.
6. Why is this the right market comparable?
An earned media value estimate asks what comparable attention might cost to buy. The word comparable does the work.
The benchmark should match the market, format, audience, time period and media quality closely enough to defend. One global CPM applied to every platform, country and content format hides rather than solves the comparison problem.
7. What changes the answer?
Some inputs are observed directly. Others, including article readership, cross-platform overlap and market rates, may be estimated. The result should therefore include a range, evidence-coverage statement and sensitivity view.
A point estimate can still appear for planning. It should never travel without the uncertainty that makes it interpretable.
A simple decision rule
Use the estimate for executive comparison only when the report can disclose all seven answers and trace every material input. Use it directionally when a limited number of gaps are explicit and bounded. Do not use it for ROI, revenue or causal impact when those outcomes have not been measured separately.
The strongest EMV report is not the one that produces the biggest number. It is the one another qualified team can challenge, reproduce and use without mistaking media equivalence for business value.
Download the 18-control EMV audit checklist, or read the full OMNI-EMV architecture.
Cite this piece
Tu Dang. "Before You Trust an Earned Media Value Number, Ask These Seven Questions." ROLearn Intelligence, August 12, 2026. https://intelligence.rolearn.dev/analysis/before-you-trust-earned-media-value-number
About the author
Tu Dang is founder of rolearn. Tu Dang is the founder of ROLearn, a market-intelligence platform used by developers, studios, and brand teams working inside virtual worlds.


