DEV Community

Oleksyj Mokrousov
Oleksyj Mokrousov

Posted on

Brand Reputation Changes From News Sentiment Shares

Negative headlines are a noisy reputation signal

Comms desks still page on “more negative articles this week.” That number mostly tracks coverage heat. Product launches and earnings weeks puff the negative lane. Wire reprints do the same. The brand’s tone mix can sit still while the pager screams - or go colder while total volume drops.

Perigon is a news and media data API. You search a large article index with company, date, and sentiment filters instead of scraping headlines by hand. This page uses that index for a portfolio diagnostic: tone shares across several large companies, then a half-month split. For a single-brand resolve → window → verify walkthrough, see Perigon’s How to Detect Brand Reputation Changes From News Sentiment.

1) Why do absolute negatives mislead a reputation desk?

2) What do tone shares look like across seven busy tickers?

3) When does a half-month split show a real mix move?

4) What do you verify before you brief a “crisis”?


Absolute negatives track volume first

On the Perigon Articles API, you pin a company (ticker or Perigon company id), set a date window, then pull three counts: all coverage, articles with strong negative sentiment, and articles with strong positive sentiment. A common cutoff is a sentiment score from 0.5 upward on each tone lane.

Brand reputation change is lane size divided by all company hits in the same window. If coverage doubles and strong negatives roughly double, the mix is stable. An absolute-count pager still fires. That is the failure mode.

Resolve the corporation first on Perigon’s Companies search. “Target” alone can return Target Australia or Target Optical before Target Corporation (NYSE TGT). “Disney” surfaces stores and studio rows before The Walt Disney Company (NYSE DIS). Save the Perigon company id; switch to it when ticker samples look off-brand.


Seven-ticker tone snapshot (30 days)

Live counts from the Perigon Articles API for publish dates 2026-08-09 through 2026-09-08. Strong negative / positive lanes use a 0.5 sentiment floor. Shares are rounded.

Ticker Company All hits Strong neg Strong pos Neg share Pos share
BA The Boeing Company ~52,200 ~24,800 ~11,400 ~48% ~22%
TGT Target Corporation ~38,700 ~15,200 ~8,200 ~39% ~21%
DAL Delta Air Lines ~12,700 ~4,800 ~2,900 ~38% ~23%
COST Costco Wholesale ~16,700 ~5,000 ~3,400 ~30% ~20%
PEP PepsiCo ~12,300 ~3,500 ~3,300 ~29% ~27%
NKE NIKE ~35,100 ~9,400 ~5,100 ~27% ~14%
DIS The Walt Disney Company ~84,300 ~21,800 ~12,500 ~26% ~15%

Boeing’s mix was the coldest here - strong negatives near half of all BA hits. PepsiCo sat near balance (~29% neg vs ~27% pos). Disney ran the largest firehose with a milder negative share than Target or Delta.

Baselines matter. Without them, week-over-week absolute spikes have nothing honest to compare against. Score ranges and average-sentiment charts sit on Perigon’s News Sentiment API guide; this page stays on share math across a watchlist.


Half-month splits: volume up, mix flat - or colder on less volume

Same company keys on the Perigon Articles API. Window A: 2026-08-09 to 2026-08-24. Window B: 2026-08-25 to 2026-09-08.

Brand Window All Strong neg Neg share
Boeing (BA) A ~27,500 ~12,200 ~44%
Boeing (BA) B ~24,700 ~12,600 ~51%
PepsiCo (PEP) A ~5,500 ~1,600 ~29%
PepsiCo (PEP) B ~6,700 ~1,900 ~29%
Delta (DAL) A ~7,500 ~3,000 ~40%
Delta (DAL) B ~5,200 ~1,800 ~34%

PepsiCo’s later half carried more coverage and more absolute negatives, yet negative share stuck at ~29%. Share detectors stay quiet while count-based alerts light up.

Boeing went the other way. Total hits fell in window B, but negatives held - share climbed from ~44% to ~51%. Colder mix on a quieter half-month. Treat that as a mix candidate, then open samples.

Delta got quieter overall. Absolute negatives fell and negative share eased from ~40% to ~34%. Milder mix, less volume - not the shape of a reputation scramble.

Keep windows equal length. Diff shares before you argue from raw counts.


Escalate only after samples survive a brand check

Open the rising tone lane and read titles. Strong-negative Boeing samples in this window included Air Canada 9/11 retrospective reprints on cp24.com and ctvnews.ca - airline-industry wires with weak Boeing aboutness - plus a Ryanair diversion on dailyrecord.co.uk that names a Boeing 737 MAX as the airframe, a product mention you should drop from a Boeing Corporation reputation brief.

PepsiCo positive-lane samples included a hospice golf fundraiser on europesays.com and a green-tea market brief on openpr.com. Cheerful copy. Thin PepsiCo attachment.

Before you brief:

1) Confirm the title is about the resolved corporation, not a supplier, rival, or airframe token.

2) Re-run the tone lane with reprints turned off when syndication is inflating the wave.

3) Drop Opinion-labeled pieces when you want reporting tone over columns.

4) Prefer the Perigon company id when ticker samples look random.

Situation Prefer
Neg count up, share flat (PEP pattern) Treat as volume change
Neg share up several points (BA pattern) Mix alert candidate - then sample
Volume and neg share both down (DAL pattern) Quieter desk
One viral reprint cluster Turn reprints off first

Key takeaways

- Absolute negatives follow coverage. Brand reputation changes from news sentiment show up as negative and positive share vs all company hits across matched windows.

- Across seven tickers on the Perigon Articles API (2026-08-09 to 2026-09-08), Boeing ran the coldest strong-negative mix (~48%); PepsiCo was near balance.

- Half-month splits: PepsiCo volume rose with flat ~29% neg share; Boeing volume fell while neg share rose ~44% to ~51%; Delta got quieter and milder.

- Verify samples and dedupe reprints before you call a tone shift a brand reputation event.

Top comments (0)