CORRECTIONS LOGGED:
ENTRIES
001. Omnicom Media Group, reporting structure
Date: February 10, 2026, 7:32 AM MST (14:32 UTC) Status: Published Subject: Omnicom Media Group, Megan Pagliuca, Joanna O'Connell What we got wrong: The piece framed OMG's product and research functions as operating on separate, disconnected tracks. The headline carried that framing. What is accurate: Joanna O'Connell reports directly to Megan Pagliuca. Omnicom Media Intelligence sits under Pagliuca's remit and she approves the research agenda. The tension the piece described is managed deliberately, not the product of two teams failing to talk. How it surfaced: OMG communications contacted us directly with the org structure. Resolution: Correction published as a standalone piece, "Plot Twist: Omnicom's Research and Product Teams Appear to Be on Speaking Terms." OMG supplied CES documentation showing research feeding product. Related: See 002.
002. Omnicom Media Group, headline framing follow-up
Date: February 10, 2026, 11:27 AM MST (18:27 UTC) Status: Published Subject: Omnicom Media Group What we got wrong: Same underlying error as 001. Logged separately because the headline was conceded as a distinct problem from the body framing, and the concession went out in correspondence with OMG the same day. What is accurate: The headline ran hotter than the reporting supported. Resolution: Acknowledged directly to OMG. Follow-up installment published focusing on OMG's product work. Note: If the public page treats 001 and 002 as one event, merge them. Kept separate here because the timestamps and the fixes were separate.
003. iSpot, layoff figures
Date: February 23, 2026, 10:37 AM MST (17:37 UTC) Status: Drafted, publication not confirmed in record Subject: iSpot.tv What we got wrong: The original series reported iSpot had cut close to 100 employees, roughly a quarter of the company. We also read the absence of WARN Act filings as evidence the company was threading a legal needle. What is accurate: Multiple independent sources subsequently described departures consistent with normal operational movement and post-acquisition reorganization following 605, not a mass reduction. On WARN, the straightforward reading is that the numbers never hit the filing thresholds. We gave the alternative framing more weight than it deserved. How it surfaced: Additional sources came forward after the original series. Resolution: Self-initiated walkback. iSpot had not responded to us at time of drafting. Action item: Confirm this ran and get the URL.
004. Marketure, room capacity
Date: March 12, 2026, 6:53 PM MST (March 13, 01:53 UTC) Draft carries stated date: March 12, 2026 Status: Drafted, publication not confirmed in record Subject: Marketure, Ari What we got wrong: We described the room as half empty. What is accurate: Marketure moved into a larger space at the venue. The room read as sparse because it was bigger, not because attendance was down. How it surfaced: Self-corrected after the fact. Resolution: Published as a correction plus clarification, with the recommendation to attend restated. Action item: Confirm this ran and get the URL.
005. Omnicom Investor Day, presenter count and gender composition
Date: March 16, 2026, 3:21 PM MST (22:21 UTC) Status: Pre-publication fix Subject: Omnicom, Investor Day What we had wrong in draft: The draft characterized the Investor Day stage as six presenters, five men and one woman, and questioned the absence of Florian Adamski and Duncan Painter. What is accurate: Nine presenters. John Wren, Daryl Simm, Phil Angelastro, George Manas, Ellen Griffin, Deepthi Prakash, Christine Gambino, Jantzen Bridges, Jacki Kelley. Five of the nine were women. Wren stated in his opening remarks that keeping CEOs off the capabilities presentations was an intentional decision, and explained why on the recording. How it surfaced: Omnicom communications answered the presenter question directly in response to our pre-publication inquiry. Resolution: Diversity section revised and the CEO-absence line held before publication. Right of reply worked as designed. Note: Caught before publication, so not a correction in the strict sense. Logged because the original framing was wrong and the record should show it.
006. Cannes analysis, subject of "The Wrong Promises"
Date: June 25, 2026, 7:22 AM MST (14:22 UTC) Status: Drafted, publication not confirmed in record Subject: Publicis, Arthur Sadoun, Omnicom, WPP What we got wrong: The Cannes piece drew a line from Sadoun's "The Wrong Promises" film to Omnicom. What is accurate: Per a holding company executive who wrote in, the film reads as a WPP mockumentary aimed at an audience of one. We did not independently confirm the 3WTC frames and attributed the claim to the source rather than asserting it. How it surfaced: Unsolicited reader correction from inside a major holding company, not for attribution. Resolution: Published as an ADOTAT+ follow-up rather than a bottom-of-page update. The correction argument concluded the misidentification made Sadoun look worse, not better. Action item: Confirm this ran and get the URL. Also flag whether the unnamed-source reliance on Publicis layoffs needs softening in the archived version.
007. The Trade Desk timing, self-flagged
Date: July 22, 2026, 7:20 AM MST (14:20 UTC) Status: Self-flagged in editorial note, not a formal correction Subject: The Trade Desk What was flagged: January call characterized in our own summary as early, with the framing that early is the polite word for wrong. Assessment: This is a self-assessment of timing, not a factual error. Logged for completeness and because the credibility framing was used publicly. Note: Probably belongs in an editorial notes column rather than corrections. Your call.
008. Agentic standards, authority attestation and decision logging
Date: July 30, 2026, 8:45 AM MST (15:45 UTC) Status: Published, correction carried inside the series Subject: IAB Tech Lab, Anthony Katsur, AAMP framework What we got wrong: Our earlier reporting argued that standardized authority attestation and durable agent decision logging sat outside every published framework in the industry. What is accurate: IAB Tech Lab's AAMP framework, per Katsur's written responses to ADOTAT, already includes buyer-side and seller-side agent reference implementations with decision telemetry and configurable guardrails. The accurate and narrower claim: the reference architecture exists inside one body, and no cross-party standard for event schema, retention, portability or audit interface has been agreed. There is still no portable audit trail a buyer, a seller and a platform can all read. How it surfaced: The correction came from the organization we were describing. Resolution: Stated first, at the top of Part II, and flagged in the series front matter. Not buried. Action item: The Part II draft carries a placeholder update-log entry pending publication date. Fill it.
009. Video Advertising Bureau, funding characterization
Date: August 4, 2026, 11:18 AM MST (18:18 UTC) Correction text carries stated date: August 4, 2026 Status: Published Subject: Video Advertising Bureau, Sean Cunningham What we got wrong: The piece described the VAB as "funded in part by Nielsen's competitors." That characterization implied a funding relationship material enough to shape VAB's conduct. What is accurate: Per VAB President and CEO Sean Cunningham, Nielsen competitors account for roughly one percent of VAB's annual revenue and hold no board seats. The board is composed of video sell-side publishers and MVPDs, who are among Nielsen's largest customers. Cunningham also states VAB has never pushed a "Nielsen is dead" narrative and that the board's position is that Nielsen should keep improving at a pace consistent with industry needs. How it surfaced: Cunningham answered a direct question. Resolution: Sentence amended in the original. Dated correction appended. Apology issued to Cunningham, the VAB, and readers. Open thread: Separately outstanding, whether Nielsen's MSA contractually prohibits side-by-side comparison or whether VAB accommodates a known preference. The Current reported the former. Do not repeat either version until it is nailed down.
RUNNING TALLY
TypeCountPost-publication corrections, published4Post-publication corrections, drafted, publication unconfirmed3Pre-publication fixes1Self-flagged editorial notes1Total events logged9Sourcing of the error, where known: Subject or subject's comms flagged it in 4 cases. Reader or source flagged it in 2. Self-caught in 2.
