AssetFlow IntelligenceResearch

Visa Inc. recorded the highest revert share at 50.00% across 8 edits

Editorial Attention · monthly · issue 2026-09-01 · data as of 2026-07-31
Which company entries in the public encyclopedia are being rewritten most heavily against their own history, and what changed?

In the 2026-07 snapshot as of 2026-07-31, Visa Inc. recorded a 50.00% revert share across 8 edits in the contested table. Eli Lilly and Company recorded a 30.00% revert share across 10 edits.

Net content reductions were recorded for Apple Inc. at -51,104 net bytes, Amazon (company) at -48,672 net bytes, and McDonald's at -5,556 net bytes.

Across editing intensity relative to each entry's own history, Eli Lilly and Company stood at the 78.20% own-history percentile. Apple Inc. reached the 74.20% own-history percentile, while Alphabet Inc. registered the 73.30% own-history percentile.

Revert shares on entries with few edits reflect small-sample editorial readings. · Net byte figures capture total length changes rather than distinguishing rewrites from deletions.
Interpretation · anthropic/claude-opus-5 · data as of 2026-07-31 · published 2026-09-01 · a view, not a cited fact; recorded so it can be scored

The heaviest rewriting this month is uncontested structural pruning of the two largest technology entries rather than any fight over what those companies are, while genuine contestation sits in a handful of low-volume entries where a single disputed sentence dominates the month. Apple Inc. shed -51,104 net bytes over 51 edits and Amazon (company) shed -48,672 over 28, yet their revert shares were only 2.00% and 3.60% — the signature of editors splitting or condensing overgrown sections with the community's assent, not of a narrative being fought over. By contrast the contested table is populated by entries with almost no volume: Visa Inc. at a 50.00% revert share on 8 edits, Nike, Inc. at 37.50% on 8, Eli Lilly and Company at 30.00% on 10 — ratios that describe two or three rejected insertions, which we read as ordinary promotional or partisan additions being patrolled away rather than sustained editorial conflict. The behavioural implication is that encyclopedia attention on mega-caps has moved from accretion to maintenance, and one-off pruning of that size is not a repeatable monthly act. In the next monthly issue neither Apple Inc. nor Amazon (company) records a net-bytes reduction larger than 10,000.

What would show this wrong: Either Apple Inc. or Amazon (company) shows net bytes below -10,000 for the 2026-08 month in the next issue.
No scoreable claim this issue: no claim menu
Checked against the tables:
  • the number '10,000' in your prose is not the value of any cited finding; cite the row (table, where, column) or remove it
  • the sentence 'The heaviest rewriting this month is uncontested structural pruning of the two largest technology entries rath' uses ['dominates', 'largest'] but no verified comparison names every row it talks about; file a finding with `relation` (and `where_b`) or a group finding whose rows are named in that sentence, or drop the comparative wording
  • the sentence 'The heaviest rewriting this month is uncontested structural pruning of the two largest technology entries rath' makes a group claim ['dominates'] with no verified group finding naming that group; file {kind: group, table, group_by: <column>, metric: count, where: {<column>: <group>}, relation: highest} or drop it
Interpretation · google/gemini-3.8-flash · data as of 2026-07-31 · published 2026-09-01 · a view, not a cited fact; recorded so it can be scored

Massive text contractions at Apple Inc. and Amazon (company) represent consensual page restructuring rather than contentious edit disputes. During the month, Apple Inc. logged a reduction of -51,104 net bytes and Amazon (company) logged -48,672 net bytes, yet both registered minimal conflict with revert shares of 2.00% and 3.60%, respectively. This stands in sharp contrast to contested entries such as Visa Inc., where edits saw a 50.00% revert share across 8 edits. Neither Apple Inc. nor Amazon (company) will appear in the contested editing table in the next monthly snapshot.

What would show this wrong: Either Apple Inc. or Amazon (company) appears in the contested editing table in the next monthly snapshot.
No scoreable claim this issue: no claim menu
Interpretation · anthropic/claude-fable-5.1 · data as of 2026-07-31 · published 2026-09-01 · a view, not a cited fact; recorded so it can be scored

The heaviest rewriting of large-company entries this month is consensual consolidation rather than contest: the two biggest content removals were carried out by many editors with almost nothing reverted, while the genuinely contested entries were small skirmishes over a few hundred bytes. Apple Inc. lost -51,104 net bytes across 51 edits by 11 human editors with a revert share of 2.00%, and Amazon (company) lost -48,672 net bytes across 28 edits by 9 editors with a revert share of 3.60%; that pattern reads as editors agreeing to trim or split sprawling pages, not as a dispute over how these companies are described. By contrast Visa Inc. topped the contested table at a 50.00% revert share on only 8 edits and a net change of 741 bytes, and no entry reached the top quintile of its own history (Eli Lilly and Company was highest at the 78.20% own-history percentile), so the month's activity is structural housekeeping of the very largest pages rather than a news-driven fight. We read this as the encyclopedia's editorial crowd treating the megacap entries as settled narratives to be tidied, which implies the removed material will stay removed. Our claim is that in the 2026-08 snapshot the Apple Inc. and Amazon (company) entries will each show net bytes below +10,000 and a revert share below 20%, confirming the removals were accepted rather than reversed.

What would show this wrong: The 2026-08 intensity table shows Apple Inc. or Amazon (company) with net bytes of +10,000 or more, or with a revert share of 20% or higher.
No scoreable claim this issue: no claim menu
Checked against the tables:
  • the number '100' in your prose is not the value of any cited finding; cite the row (table, where, column) or remove it
  • the number '+10,000' in your prose is not the value of any cited finding; cite the row (table, where, column) or remove it
  • the number '20%' in your prose is not the value of any cited finding; cite the row (table, where, column) or remove it
  • the sentence 'By contrast Visa Inc. topped the contested table at a 50.00% revert share on only 8 edits and a net change of ' uses ['highest', 'largest', 'top'] but no verified comparison names every row it talks about; file a finding with `relation` (and `where_b`) or a group finding whose rows are named in that sentence, or drop the comparative wording

Editing intensity in the latest full month, against each entry's own history

NameEntryEditsHuman editors× own medianOwn-history percentileRevert shareNet bytes
LLYEli Lilly and Company1052.8678.20%30.00%525
AAPLApple Inc.51111.8574.20%2.00%-51,104
GOOGAlphabet Inc.25101.4773.30%12.00%319
MCDMcDonald's22131.9172.70%13.60%-5,556
MAMastercard1251.5071.40%8.30%-719
METAMeta Platforms21121.6869.90%4.80%1,328
NFLXNetflix, Inc.981.1369.20%0.00%407
NVDANvidia18110.5163.30%11.10%2,439
HOODRobinhood Markets531.6761.10%0.00%351
PGProcter & Gamble1241.7159.60%0.00%17
VVisa Inc.861.4558.30%50.00%741
TSNTyson Foods532.5055.90%0.00%-1,238
MSFTMicrosoft18131.1355.50%0.00%3,975
NKENike, Inc.842.0055.00%37.50%461
AMZNAmazon (company)2892.1548.10%3.60%-48,672
AMDAMD1160.8847.10%0.00%742
CSCOCisco1071.8245.00%0.00%602
PEPPepsiCo731.7544.30%0.00%3,561
HDHome Depot981.2942.60%11.10%258
DISThe Walt Disney Company1671.2838.00%0.00%612
PFEPfizer661.2034.80%0.00%209
WMTWalmart20101.3334.60%5.00%3,392
TAT&T15101.0334.20%6.70%197
JNJJohnson & Johnson540.6326.70%20.00%-300
XOMExxonMobil540.7125.30%0.00%-431
KOThe Coca-Cola Company880.7622.40%12.50%-91
BRK.ABerkshire Hathaway530.5020.60%20.00%471
TSLATesla, Inc.1480.7419.30%28.60%196
SBUXStarbucks760.7017.80%0.00%1,877
INTCIntel1060.3416.70%10.00%751
Each entry is the company's own public-encyclopedia page, named explicitly in the watch list and measured against its own monthly history (raw edit counts are never compared across entries — editing peaked site-wide in 2007). Subsidiary, product and brand pages are not included. A high revert share means contested editing, not vandalism per se. Entries with fewer than five edits in the month are omitted. Month: 2026-07 (all rows).

Contested editing in the latest full month (share of the month's edits that were reverted)

NameEntryEditsHuman editorsRevert shareNet bytes
VVisa Inc.8650.00%741
NKENike, Inc.8437.50%461
LLYEli Lilly and Company10530.00%525
TSLATesla, Inc.14828.60%196
BRK.ABerkshire Hathaway5320.00%471
JNJJohnson & Johnson5420.00%-300
Contested = the share of the month's edits to the entry that were reverted; shown only where at least a fifth of the month's edits were reverted. A revert share on a handful of edits is a small-sample reading.

Largest net changes in the month

NameEntryNet bytes|Net bytes|Edits
AAPLApple Inc.-51,10451,10451
AMZNAmazon (company)-48,67248,67228
MCDMcDonald's-5,5565,55622
MSFTMicrosoft3,9753,97518
PEPPepsiCo3,5613,5617
WMTWalmart3,3923,39220
NVDANvidia2,4392,43918
SBUXStarbucks1,8771,8777
A large negative net-bytes figure means content was removed; whether that was a rewrite, a split or a strip is in the edit summaries.

Substantive changes on entries in the top quintile of their own editing history (verbatim edit summaries)

Snapshot

History snapshotLast full monthBuilt
2026-072026-072026-08-25
Edit history comes from a pinned monthly snapshot; months after the last full month are not covered.

Ask deeper

What changed on Visa_Inc.? →Eli_Lilly_and_Company: edits vs attention →Contested entries →

Method

Each month the run reads the editing history of the public-encyclopedia entries for a fixed watch list of 46 large companies — each named explicitly as a (ticker, page title) pair in the recipe, so the same 46 entries are measured every month — and measures the latest full month against each entry's own history: edits, distinct human editors, the ratio to the entry's median month, the own-history percentile, the share of edits that were reverted, and the net change in bytes. The contested table lists the entries where at least a fifth of the month's edits were reverted. For the entries in the top quintile of their own history the run pulls the substantive revisions and reproduces the editors' own summaries, with housekeeping edits set aside by an advisory classifier.

Every entry is the company's own page (Apple_Inc., Procter_&_Gamble, Walmart, ...), pinned by title. Subsidiary, product and brand pages that the estate also links to a ticker are deliberately left out, so an entry's editing is that company's editing and nothing else. Two limits follow from the pin: a page renamed in the the encyclopedia after the pinned snapshot keeps its old title here, and a company whose own page saw fewer than five edits in the month does not appear. Net bytes is the month's total change in the entry's length: a large negative figure means content was removed, and only the edit summaries say whether that was a rewrite, a split or a strip.

The history comes from a pinned monthly snapshot, so an issue publishes only when the covered month has moved, and the issue's data date is the last day of that month. Editing intensity is an editorial measure of how much a public record is being rewritten. It is not a valuation and carries no return claim.

Because the history is a pinned snapshot that does not refresh between issues, the interpretation on this page carries no scored claim: nothing a later snapshot could resolve it against exists until the store is deliberately re-pinned, so the view is published as a view and the page says so rather than logging a claim that cannot be scored.

Published by AssetFlow.ai LLC. Descriptive research on attention and public-record data; not investment advice and not an offer or solicitation. No return claim attaches to anything above. Issue date 2026-09-01.