Key takeaways
- Share of voice tells you how loud you were last quarter. It does not tell you what anyone now believes.
- Volume reporting collapses eight distinct questions into a single ratio, which is why coverage totals and executive confidence so often move in opposite directions.
- A working measurement taxonomy separates concentration, prominence, message pull-through, source diversity, velocity, competitive framing, stakeholder relevance, and evidence gaps.
- Each measure answers a different business question, so the right metric depends on the decision in front of you rather than on whatever the dashboard defaults to.
- AI systems now read the same coverage your stakeholders read, which turns source quality and evidence gaps into operational concerns.
- Pick three measures tied to live decisions this quarter and retire anything that only produces a chart.
Most communications reports open the same way. Total mentions, then share of voice, then a sentiment split, then a slide of logos. Those numbers describe the volume of a conversation without describing its content, and that gap is where the useful signal lives. The media intelligence metrics that actually change decisions measure which stories are forming, who is being positioned to own them, and which claims are being repeated by sources that carry weight. Reputation has also become more volatile than any single reading can capture.
Axios reported in May 2026 that more than half of the companies in the Axios Harris Poll 100 that also appeared in the prior year's rankings saw their reputations improve year over year, a rebound the outlet attributes in part to fragmented news consumption that lets audiences move past scandals faster. When reputation moves that quickly in both directions, a quarterly snapshot says very little about direction. Reading narratives instead of tallies is what a modern communications intelligence approach is designed around.
What Are Media Intelligence Metrics Beyond Share of Voice?
Beyond share of voice, the measures that matter describe the shape and quality of coverage rather than its volume. They answer questions a percentage cannot, because a percentage can only report how much of a conversation you occupied, never what that conversation established about you. Volume answers what happened. The rest of the taxonomy answers whether it mattered and what to do about it.
The difference shows up in the meeting. A review built on volume produces a discussion about whether the number went up. A review built on narrative structure produces a discussion about which story lines to reinforce, which to reframe, and where the supporting evidence is too thin to hold. That second conversation is the one senior leadership was asking for.
Why Is Share of Voice Alone a Weak Basis for Decisions?
Share of voice is a ratio, and ratios hide their inputs. Two companies can hold an identical share of a category conversation while occupying opposite strategic positions: one appearing prominently in the three outlets its investors read, the other appearing in passing across hundreds of syndicated reposts. Share of voice alternatives are not replacements for the metric. They are the context that makes it interpretable. A single share figure conceals three things worth separating:
Whether the coverage appeared in outlets your audience actually reads
Whether your brand was central to those stories or incidental to them
Whether the story lines behind the coverage are gaining or losing ground
None of that is visible in the ratio itself. Surfacing it takes measures that describe composition rather than size, which is why rate of change and narrative structure carry more decision value than any single percentage. That reframing sits at the center of modern PR measurement standards.
Which Metrics Should Communications Leaders Track?
Eight measures cover the ground that volume reporting leaves open: narrative concentration, prominence, message pull-through, source diversity, narrative velocity, competitive framing, stakeholder relevance, and evidence gaps. A taxonomy of media intelligence metrics works only when each entry earns its place by answering a question no other entry answers. The eight below are deliberately non-overlapping, and no team needs all of them running at once. Several of them are narrative intelligence metrics rather than coverage metrics, meaning they describe story lines instead of counting articles. Treat this as a menu matched to decisions rather than a scorecard to fill in.
Narrative Concentration
Concentration measures how much of your coverage is carried by a small number of story lines versus scattered across many. High concentration means a handful of ideas now define you, for better or worse. Low concentration means your reputation is diffuse and nothing has consolidated, which is a different problem requiring a different response.
Where concentration runs that high, those three story lines are effectively your public identity, and everything else in the coverage is noise around them.
Prominence
Prominence measures where the brand sits inside the coverage: the subject of the piece, prominently featured, meaningfully discussed, or mentioned once in a list. Volume tools count all four identically, which is how a team reports record coverage while leadership feels invisible. Tracking prominence separately makes the difference between being the story and being background detail visible in a way a mention count never will.
Message Pull-Through
Message pull-through asks whether the idea you set out to establish actually appeared, and it should never be scored as a simple yes or no. The useful version measures how much of the coverage carried the message, how much of that landed in outlets your audience trusts, and whether the brand was prominent when it did. It also asks an uncomfortable question worth asking: which competitor is currently attached to the same idea?
Source Diversity
Source diversity measures how many genuinely independent outlets support a given claim about your company. Twenty URLs originating from one wire release represent one editorial decision replicated twenty times, not twenty confirmations. This distinction has become more consequential as AI systems weigh evidence rather than repetition, a dynamic that turns on which sources AI systems trust.
Narrative Velocity
Velocity measures rate of change rather than level. A negative story line sitting at low volume but accelerating across authoritative publications is more urgent than a stable, high-volume, neutral one. Because velocity is a derivative, it surfaces a developing situation days or weeks before a volume chart registers anything unusual. That lead time is the entire value of the measure.
Competitive Framing
Competitive framing is not a count of how much coverage rivals received. It measures how each company is positioned inside the same story lines: which one is cast as the category leader, which as the fast follower, which as the cautionary example. A company can win share of voice in a category and still lose the framing battle inside every article, and only framing analysis will show it.
Stakeholder Relevance
Stakeholder relevance weights coverage against the audiences whose beliefs affect the business. A trade publication read closely by two dozen regulators can outweigh a national feature read by two million people who will never make a decision about your company. This is the measure that reconciles communications measurement with how the business defines its priorities, and it is the one most often missing.
Evidence Gaps
Evidence gaps measure negative space. For each strategic claim the organization wants to be believed, the metric asks whether independent, current, specific supporting evidence exists in the public record. Most measurement programs only count what is present, which means the most actionable finding in a reputation review is frequently the one nobody is tracking: the message leadership considers central, with almost nothing credible behind it.
How Do You Match a Metric to the Business Question?
Start with the decision on the table, then pick the measure that informs it. A named set of media intelligence metrics only pays off at this step, because it lets you work backward from a question to the evidence rather than opening a dashboard and hoping a decision emerges. The table below maps the eight measures to the questions they can actually answer.
| Metric | What it measures | When to use it | Business question it answers |
|---|---|---|---|
| Narrative concentration | Share of coverage carried by top story lines | Annual positioning reviews, post-campaign | What are we currently known for? |
| Prominence | Brand centrality within each article | Ongoing quality control on coverage | Are we the story or the backdrop? |
| Message pull-through | Reach and placement of intended messaging | Campaign and launch evaluation | Did our message take hold where it counts? |
| Source diversity | Independent outlets supporting a claim | Claim validation, crisis assessment | Is this widely believed or widely copied? |
| Narrative velocity | Rate of change in a story line | Continuous risk and opportunity watch | What is forming that we should act on now? |
| Competitive framing | Positioning of rivals within shared narratives | Category and market-position reviews | Who is winning the story, not the volume? |
| Stakeholder relevance | Coverage weighted by audience importance | Board and executive reporting | Did the people who matter see this? |
| Evidence gaps | Missing support for strategic claims | Pre-campaign planning, annual audits | What do we assert that we cannot prove? |
How Often Should Each Metric Be Reviewed?
Cadence matters as much as selection, because a measure reviewed on the wrong rhythm produces either noise or hindsight. Velocity read quarterly is a post-mortem. Concentration read weekly is static.
Recommended starting cadence. Adjust within one quarter based on how fast your category moves.
| Cadence | Metrics best suited | Purpose |
|---|---|---|
| Weekly | Narrative velocity, prominence | Catch developing story lines early enough to act |
| Monthly | Message pull-through, competitive framing, source diversity | Assess whether campaigns are landing as intended |
| Quarterly | Narrative concentration, stakeholder relevance, evidence gaps | Reset positioning and identify what needs building |
What Changes When AI Systems Read the Same Coverage?
Earned media now reaches stakeholders through two paths. People still read articles directly. They also ask an AI assistant to explain a company, compare it to competitors, or summarize why its reputation is shifting, and receive a synthesized answer built from that same coverage. Large language models are now a genuine audience for communications work, one that forms and repeats conclusions about your brand without ever visiting your website.
That audience is also becoming a trusted one. The Capgemini Research Institute's 2026 consumer trends report, based on a survey of 12,000 people across 12 countries, describes AI moving from enabler to trusted advisor, and found that two-thirds of respondents trust AI more when it explains the reasoning behind its recommendations. That detail matters for communications. A synthesized answer becomes more persuasive when it can show its evidence, which means the evidence available about your company is doing more work than it used to.
Two entries in the taxonomy carry most of the weight here. Source diversity determines whether a favorable claim about your company is supported by independent evidence or by one release copied across the web, and independent evidence is what survives synthesis. Evidence gaps determine whether an AI system can find anything credible to support your central positioning at all, and an absence of evidence tends to resolve as an absence of the claim. Measures that stop at volume cannot see either condition, and neither can PR measurement metrics built only around reach. That is why counting appearances in AI answers is a thin substitute for understanding the evidence behind them.
Frequently asked questions
Three is the right starting number and five is the practical ceiling, each tied to a specific decision on the calendar. Teams that instrument all eight simultaneously tend to produce comprehensive reports nobody acts on, which recreates the problem the taxonomy was meant to solve.
No. Share of voice remains a useful competitive baseline, particularly when segmented by publication tier, prominence, and specific messages. The failure mode is treating it as a verdict rather than a starting point.
Stakeholder relevance and narrative concentration translate most cleanly for executive audiences, because both connect directly to strategy. Concentration answers what the company is known for. Relevance answers whether the audiences that determine outcomes actually encountered it.
Article sentiment scores the emotional tone of a piece of coverage. Narrative sentiment scores how the brand is positioned within the story line, which can diverge sharply. A story about industry turmoil may read negatively while positioning your company as the credible authority explaining it.
Choose the Three Metrics Your Next Board Update Actually Needs
Choosing media intelligence metrics well comes down to knowing which question you are answering and which measure answers it, then building the reporting rhythm around live decisions instead of reporting cycles. Start with the decisions your leadership team will make this quarter, select the three that inform them, and treat everything else as optional.
Handraise was built for exactly this: clustering coverage into the narratives that define a brand, tracking how those narratives move across both media and AI systems, and recommending the messaging most likely to shift how both humans and AI systems describe you. Book a briefing to see what your current narrative picture looks like.