Narrative Intelligence for Investor Relations

Read the narrative around your company.

Not a sentiment score. The interplay between what management says, how sell-side analysts interpret and challenge it, and how the financial press reframes it. Tracked quarter over quarter so every earnings cycle sharpens the next.

3 voices · Management · Sell-side · Press Quarter over quarter Pre & post-earnings Peer-benchmarked
CEO Narrative Leaderboard · Enterprise Software
Q2 2026 · trailing 4 quarters · illustrative
# CEO Company NarrativeScore Archetype Consistency
1Satya NadellaMicrosoft74 +5VISIONARY89
2Olivier Le PeuchSLB N.V.72 +2VISIONARY82
3Joaquin DuatoJ&J71 +3VISIONARY82
4James GormanMorgan Stanley71 +1STEWARD88
5Safra CatzOracle71 −1COMMANDER85
How to read. NarrativeScore (0–100) measures how clearly management's intended story is landing with sell-side and press, with the delta vs. prior quarter. Archetype captures the dominant narrative posture. Consistency is the cross-stakeholder agreement on what the company stands for.
The science · the tensor

Narratives are not flat. They are read differently by every stakeholder, on every horizon.

Traditional sentiment reduces a story to a single score, one reader, one moment. Pharos models how the same narrative is interpreted across stakeholders, time horizons, and signal types. The latent structure beneath what markets actually price.

Conventional sentiment
One score. One story. One reader.
A single polarity number, averaged across whatever was published. Useful for trend lines, useless for action. Tells you the temperature, not what's actually moving or who's moving it.
flat averaged stakeholder-blind
The Pharos tensor
Multidimensional. Stakeholder-aware. Time-resolved.
Every narrative is decomposed across three axes: who's interpreting it, on what horizon, and what type of signal it carries. The output is a structured read on where the conversation actually is, not a number averaged from where it has been.
stakeholder axis time-horizon axis signal-type axis
The IR workflow

One lens, applied twice. Before the call to prepare. After the call to learn.

Two connected modules of the same underlying tensor. One readies the room before management steps on stage. The other closes the loop on how the story actually moved, so the next quarter starts with sharper inputs, not the same blind spots.

Before the call

SpeechScape

Pre-earnings · executive preparation
Walk into the call knowing who is in the room, what they're likely to ask, where the narrative has drifted from your intended message, and which executive each question should route to.
  • Predicted analyst attendance & questioning roster
  • Questioning DNA per analyst: skepticism, specificity, follow-up rate
  • CEO ↔ CFO routing: which exec answers which question better
  • Likely pressure themes, ranked by probability
  • Relationship drift signals: strengthening and declining
  • Peer benchmarking: how your narrative ranks in category
After the call

NewsScape

Post-earnings · narrative monitoring
Watch how the story actually moves in the days that follow. Which themes landed, which got reframed, how analyst stance shifted, and where the press contradicts what management said.
  • Press interpretation, theme by theme
  • Management vs. press divergence, scored daily
  • Contradiction detection: what the press is pushing back on
  • Credibility & novelty shifts vs. prior quarter
  • Contagion across peers: what's spreading and where
  • Stakeholder-aware interpretation, not aggregate sentiment
What "moved" actually looks like

Three narratives, one lens. Scored, compared, and ranked.

Management framing. Press framing. Peer-press framing. Three voices on the same earnings story, tracked from the moment the call ends through the days that follow.

Narrative Divergence · Mgmt vs Press vs Peer Press · Q2 2026 · illustrative
14 sources scored · post-earnings · 9-day window
+60 +30 0 −30 CALL ENDS +2D +4D +6D +9D MGMT PEER PRESS
Management framing
Press framing
Peer-press framing
Divergence · Mgmt vs Press
+58.2
Credibility · vs prior call
0.62
Novelty · vs peer avg
0.71
Contradictions flagged
3
Why this is different

Conventional IR tools count what already happened. Pharos explains what's moving, where, and what to do about it.

Vs. sentiment scoring

Sentiment averages noise. The tensor reads structure.

One number averaged across whatever was published tells you the weather, not the system underneath it. We model who is reading the story, on what horizon, and with what kind of signal, so you can act on the divergence, not just track it.

Vs. media monitoring

Monitoring is a feed. This is a read.

Clip alerts and PR dashboards tell you what was published. Pharos tells you how the narrative moved. Which themes the press reinforced, which it reframed, where it contradicted management, and what's spreading across the peer set.

Vs. analyst Q&A prep memos

Prep memos rely on memory. We score every call.

Most IR teams prepare by re-reading last quarter's transcript and guessing who'll be on. We score every analyst across every S&P 500 call they touch. You get the questioning DNA, not the vibe.

Source-agnostic engine

One engine. Many sources. Every signal that shapes your narrative is fair game.

The same tensor reads structured and unstructured signal alike, from the call itself to the filings around it, from sell-side notes to the financial press to the lobbying record.

Earnings call transcripts
Mgmt & Q&A · scored per exchange · scoped per analyst.
Sell-side analyst reports
Consensus narrative. Who's echoing, who's pushing back.
Financial press
Coverage interpretation, contradiction detection, contagion across peers.
SEC filings & investor events
10-K · 10-Q · 8-K · Investor Day. Cross-reference call narrative vs. disclosure.
Get the read

Request a narrative read on your last earnings cycle.

Send us your ticker. We'll come back with the pre-call analyst roster, the management-vs-press divergence on the themes that mattered, and a short read on where the narrative drifted. Built on the same engine that powers the workflow.