Kscope · deep dive · 2026-10-14
Diligent sells a board portal and a market-intelligence file of activism, votes, and pay. The portal runs on documents the client uploads. The US ballot, the fund vote, and the pay table are filings. N-PX was loaded on 2026-10-14.
The row stays now. N-PX is 72,435,308 votes on 33,944 filings, every filer, latest filing 2026-10-13, refreshing daily. The structured form starts with reporting periods ending 2024-06-30. Proposals extracted from the proxy are 125,520. Vote results from the 8-K are 188,462. Contested-proxy events are 11,127 on 1,206 companies, from 2016 through 2026-09-18, and most activist names are still unresolved. Pay, as of 2026-09-19, is 653,295 distinct person-company-years. Their pages call the voting and activism files the largest of their kind and do not publish a count to set beside these. The board portal and the governance suite stay theirs. A vote or campaign outside a US filing stays theirs.
| Feature | Diligent | Kscope | |
|---|---|---|---|
| Proposals on the proxy | Diligent Market Intelligence sells shareholder activism, proxy voting, compensation, and ESG. The product page calls the voting file the largest of its kind and says coverage is global. The pages read on 2026-10-14 do not publish a count of proposals or meetings. | 125,520 proposals extracted from definitive proxies. Meeting dates come from the proxy cover and the nearest vote result. 2,461 rows still have no meeting date. A US proposal on the proxy is in the file. | Ahead |
| How the meeting voted | Vote results sit in the same voting product, with investor-level patterns and the rationale a voter publishes. | 188,462 vote-result rows extracted from 8-K item 5.07. 1,800 of them still have no meeting date. The result of a US meeting is the filing. | Ahead |
| N-PX fund votes | The market-intelligence page offers a trial on live 2026 N-PX data. They describe a global voting archive and do not publish a filing count or a vote count. | Loaded 2026-10-14 and refreshing daily. 33,944 filings, the latest filed 2026-10-13. 35,725 voters, 2,473,249 ballot items, and 72,435,308 votes. Every filer is in the file. The structured XML starts with reporting periods ending 2024-06-30. 585,744 items are tied to the issuer's proposal, about 93 percent of the items whose meeting is already in the proxy file. Four Dimensional filings stop at 500,000 vote rows inside the source document and are marked incomplete. A manager rollup is on the API. BlackRock's 2026 file is 1,837,082 fund votes, 7.04 percent against management. | Ahead |
| Proxy-advisor recommendations | The product page sells investor rationale and voting-policy data. The page read on 2026-10-14 does not name ISS or Glass Lewis. Older data-feed copy listed both firms' recommendations. | A recommendation trained on the ballot and the votes is our model. A recommendation that exists only because ISS or Glass Lewis wrote it stays a license. | Beside |
| Voting policies | Investor policies and the rationale behind a vote are part of the stewardship feed. A buyer can apply a custom policy across the book. | A policy a fund publishes is a crawl. A policy that exists only in their file stays theirs. Running our own policy across the N-PX book is a model on votes we hold. | Partial |
| Feature | Diligent | Kscope | |
|---|---|---|---|
| Contested proxies | Real-time and historical activist campaigns, with global coverage and alerts. They call the activism file the largest of its kind. The pages read do not publish a count of campaigns, companies, or markets. | 11,127 events on 1,206 companies, from 2016-01-04 through 2026-09-18, taken from the contested-proxy and tender-response forms. 298 of 2,530 activist names resolve to an investor profile. Campaign language inside an ordinary 10-K, proxy, or 8-K is not a second extract yet. A demand that never hits a US filing stays theirs. | Partial |
| Vulnerability score | They rank how exposed an issuer is to an activist. The score is theirs. | A score trained on the filings, the votes, and the campaign events is our model. | Partial |
| Pay | Structured executive compensation, and a pay plan that stands up to scrutiny. The page read on 2026-10-14 does not publish a company count. | Pay is extracted from the definitive proxy. As of 2026-09-19 the distinct cut is 653,295 person-company-years, 619,494 after dropping totals that fail a sanity check, across 8,076 issuers and 114,858 people. Two extractors overlap, so a raw row count is not the coverage. A third parser, reading the HTML table itself, is in progress, and the headline can move when it lands. The year field runs out to 2029 on a handful of rows, which is a parse artifact. Director pay and audit fees are extracted from the same proxies. A pay-for-performance score we train is our model. | Match |
| Feature | Diligent | Kscope | |
|---|---|---|---|
| Board portal | Diligent Boards, BoardEffect, and GovernAI run the meeting: the board book, the minutes, and an assistant over the client's own materials. GovernAI was rewritten in a note dated 2026-10-14. | Those documents exist because the client uploaded them. The portal is their software. | Gap |
| Entity records and the GRC suite | Entities keeps the subsidiary register the client maintains. The same platform sells risk, audit, policy, whistleblowing, third-party diligence, and compliance training. | A subsidiary named in a filing is on the company graph. The register a secretary maintains inside their tool, and the rest of that software suite, stay theirs. | Gap |
| What a seat costs | Market Intelligence is request-a-demo. The pages read publish no dollar seat price. A trial is offered against the 2026 N-PX file. | The US ballot, the fund vote, and the pay table are already in house. A seat price for this screen is not set. | Ahead |
Ahead The feature is data we already hold, in a form an investor can use.
Match Both products do this.
Partial We have the disclosed half, or the raw material for the model, and not their finished file.
Gap They ship it. We do not.
Beside The ISS or Glass Lewis recommendation is a license. A recommendation we train is our model.