AI-generated
Moatsight's own AI workflow wrote this edition. No person edited it.
- Written Nov 10, 2026 from 25 sources, including the 10-Q filed Nov 9, 2026.
- Argued by a bear and a bull critic.
- Before it was published, a separate AI check traced the key figures the company reported to saved copies of its filings.
Earnings
Loud claims, checked: 1 not supported · 3 partly supported · 3 open
| Verdict | Claim | Who and when |
|---|---|---|
| Partly supportedThe word opted-in was dropped from the 10-K after fiscal 2023, and the company's November 2024 counts give 110 million US email opt-ins against 245 million with digital permission; whether investors were misled is open | Zeta's US data set was not opted-in as the company once described itIn Business explainer | Culper Research, short report of November 13, 2024; plaintiffs in the securities class action |
| OpenThe company says revenue from customers who are also vendors is insignificant; no court, regulator or restatement has decided the claim | Revenue was flattered by two-way contracts with data suppliers, and much of the data came from consumer websites that do not make its use clearIn Business explainer | Culper Research, November 13, 2024 |
| OpenAn allegation in a complaint that no court has decided | The chief executive sold shares through trusts and companies to avoid reportingIn Business explainer | Plaintiffs' amended complaint in the securities class action, May 2025, as PPC Land reported on July 13, 2026 |
| Not supportedOn the present filings: no AI or partnership revenue is reported and revenue is still marketing revenue | Zeta is now an intelligent AI infrastructure company, comparable with Palantir, Datadog or SnowflakeIn What changed | The company, in its release and earnings supplemental of November 9, 2026 |
| OpenNo partnership revenue is disclosed and no year is given | The Palantir partnership can add more than $100 million of annual revenueIn What changed | The CEO, on the November 9, 2026 earnings call |
| Partly supportedOrganic growth rose from 28% to 30%; most of the rest came from Marigold and election spending | Revenue growth of 46% shows the business acceleratingIn What changed | The company's release of November 9, 2026 |
| Partly supportedThe beats are real but run at about the 2% to 5% cushion the CFO says is built into guidance | Twenty-one straight beat-and-raise quarters show momentum beyond planIn What changed | The company's release of November 9, 2026 |
A verdict is the Moatsight view of what the filings show, not a finding of fact.
The core#
The third quarter of fiscal 2026 (to September 30, 2026) was the best Zeta has reported. Revenue of $492.7 million grew 46%, about 30% without the Marigold acquisition and election spending. Adjusted EBITDA reached a 24.6% margin, GAAP net income was $38.3 million, and free cash flow as the company defines it was $81.5 million.
The headline hides three things:
- A third of the growth was bought.
- Stock-based compensation of $50.0 million still takes most of the cash a shareholder would count. Free cash flow less stock-based pay was $31.5 million in the quarter and only $38.2 million over twelve months.
- The guided fourth quarter implies organic growth slowing to 22–24%, though the company's own record suggests the guide is cushioned.
On the management record, the word is reliable on the numbers it guides. Every quarterly and full-year figure in the record since late 2024 came in above the top of its range. The company also openly builds a 2–5% cushion into its outlook. That is a strength, but it also means a "raise" is mostly the release of a cushion that was always there.
The quarter against guidance and a year ago#
| Metric, third quarter | Q3 2026 | Q3 2025 | Change | Guidance given August 4, 2026 | Against guidance |
|---|---|---|---|---|---|
| Revenue, USD million | 492.7 | 337.2 | +46.1% | 469 to 472 | 4.4% above the top |
| Revenue excl. Marigold and political candidates, USD million | 437 | 337 | +30% | not guided in dollars; growth of 23% to 24% "excluding M&A and political" | about 6 points above |
| Adjusted EBITDA, USD million | 121.0 | 78.1 | +55.1% | 115.0 to 116.0 | 4.3% above the top |
| Adjusted EBITDA margin | 24.6% | 23.2% | +1.4 points | 24.4% to 24.7% | inside the range |
| GAAP operating income, USD million | 45.9 | 8.8 | +425% | not guided | |
| GAAP net income, USD million | 38.3 | -3.6 | n/m | not guided | |
| GAAP diluted EPS, USD per share | 0.14 | -0.02 | n/m | not guided | |
| Free cash flow (company definition), USD million | 81.5 | 47.1 | +73.0% | not guided for the quarter | |
| Stock-based compensation, USD million | 50.0 | 45.6 | +9.5% | not guided |
n/m means not meaningful, because the year-earlier figure was a loss. The revenue rows excluding Marigold and political candidates are the company's own rounded figures. The 30% is our calculation of $437 million against $337 million. The company also states it as "30% Y/Y".
How to read this: The company's reconciliation table removes LiveIntent revenue ($20.6 million) from the third quarter of 2025, leaving $317 million. Dividing $437 million by $317 million gives 38%, which is not the right rate. LiveIntent, bought in October 2024, has been part of both years' revenue since the fourth quarter of 2025. The company's 30% compares $437 million with the full $337 million, which is like for like. The same logic gives the fourth-quarter guide of "22% to 24%": $463 million of guided organic revenue against $376 million in the fourth quarter of 2025 without Marigold.
What the headline hides#
Growth: bought, seasonal and organic#
The third quarter's revenue rose by $155.5 million. The 10-Q splits the increase into $87.2 million from new customers and $68.3 million from existing ones. Inside the new-customer figure sits Marigold's $48.0 million, so new organic customers added about $39 million, which we calculate. Political candidates added $7.4 million, against an amount the company calls "not material" a year earlier. Taking both out, organic revenue grew about 30%. That was the third quarter running at 28–30%, and it is the rate that matters for value.
| Quarter | Revenue, USD million | Growth Y/Y | Growth excl. acquisitions and political (company), Y/Y | Adjusted EBITDA margin | SBC, % of revenue | GAAP operating margin | GAAP net income, USD million |
|---|---|---|---|---|---|---|---|
| Q4 2024 | 314.7 | n/a | 31% | 22.4% | 13.7% | 2.2% | 15.2 |
| Q1 2025 | 264.4 | 35.6% | 25.9% | 17.7% | 15.9% | -6.1% | -21.6 |
| Q2 2025 | 308.4 | 35.4% | 27.4% | 19.1% | 15.1% | -1.7% | -12.8 |
| Q3 2025 | 337.2 | 25.7% | 28% | 23.2% | 13.5% | 2.6% | -3.6 |
| Q4 2025 | 394.6 | 25.4% | 28% | 24.1% | 11.1% | 4.5% | 6.5 |
| Q1 2026 | 396.3 | 49.9% | 29% | 16.7% | 13.4% | -4.8% | -13.2 |
| Q2 2026 | 442.8 | 43.5% | 28% | 20.7% | 11.8% | 3.8% | 8.2 |
| Q3 2026 | 492.7 | 46.1% | 30% | 24.6% | 10.1% | 9.3% | 38.3 |
Notes on the table:
- Growth for the fourth quarter of 2024 is n/a because the series starts there.
- The organic rates for the first two quarters of 2025 are the ones the CFO gave on the February 24, 2026 call.
- Organic rates for 2025 exclude LiveIntent; those for 2026 exclude Marigold.
The table shows a strong seasonal pattern. The first quarter is the weakest each year, in both revenue and margin, and the fourth quarter is the strongest. Comparisons should therefore be made against the same quarter a year earlier, never against the quarter before.
Margins: the adjusted figure and what the shareholder keeps#
Adjusted EBITDA leaves out stock-based compensation. Most of Zeta's progress is real nonetheless, because stock pay has grown much more slowly than revenue. The incremental margin is the change in profit divided by the change in revenue against the same quarter a year earlier. We calculate it on two bases:
| Quarter | Incremental adjusted EBITDA margin, Y/Y | Incremental margin of adjusted EBITDA less SBC, Y/Y | Incremental GAAP operating margin, Y/Y |
|---|---|---|---|
| Q4 2025 | 30.9% | 30.0% | 13.7% |
| Q1 2026 | 14.7% | 6.4% | -2.1% |
| Q2 2026 | 24.5% | 20.3% | 16.4% |
| Q3 2026 | 27.6% | 24.9% | 23.9% |
In the third quarter, each extra dollar of revenue brought 24 cents of GAAP operating profit. That is the first quarter where the GAAP figure comes close to the adjusted one. The weak first quarter of 2026 carried Marigold's integration: $6.8 million of restructuring, and research and development up from $26.8 million to $45.0 million.
GAAP net income also benefited from a smaller charge for remeasuring acquisition payments. Other expenses fell from $11.7 million to $2.7 million, a swing of $9.0 million. Operating income, at $45.9 million against $8.8 million, shows that the improvement does not depend on that swing.
The definition of the company's "non-GAAP net income" changed in the third quarter, as the third-quarter 2026 supplemental notes. It now adds back amortisation of acquired intangibles ($13.6 million in the quarter), with prior periods recast. The non-GAAP EPS of $0.38 per share is therefore not comparable with figures the company published before November 2026. The GAAP diluted EPS of $0.14 per share is the figure we use.
Cash: conversion is good, the cost of stock pay is not#
| Quarter | Free cash flow (company), USD million | FCF margin | Stock-based compensation, USD million | FCF less SBC, USD million | FCF less SBC, % of revenue |
|---|---|---|---|---|---|
| Q4 2024 | 31.7 | 10.1% | 43.0 | -11.3 | -3.6% |
| Q1 2025 | 28.2 | 10.7% | 42.0 | -13.8 | -5.2% |
| Q2 2025 | 33.6 | 10.9% | 46.5 | -12.9 | -4.2% |
| Q3 2025 | 47.1 | 14.0% | 45.6 | 1.5 | 0.4% |
| Q4 2025 | 55.8 | 14.2% | 43.7 | 12.1 | 3.1% |
| Q1 2026 | 41.7 | 10.5% | 53.0 | -11.4 | -2.9% |
| Q2 2026 | 58.0 | 13.1% | 52.1 | 5.8 | 1.3% |
| Q3 2026 | 81.5 | 16.5% | 50.0 | 31.5 | 6.4% |
Free cash flow converted 67.3% of adjusted EBITDA in the third quarter, against 60.4% a year earlier. The company's definition subtracts capitalised software as well as capital expenditure, which is stricter than many peers' definitions.
Over the twelve months to September 2026:
- Free cash flow was $237.0 million.
- Stock-based compensation was $198.9 million.
- FCF less SBC was therefore $38.2 million, 2.2% of revenue.
- The third quarter alone produced $31.5 million of that.
Stock pay is an expense paid in shares rather than cash. A shareholder bears it as dilution, or as the cash spent buying the shares back. The trend is right: stock-based compensation fell from 15.9% of revenue in the first quarter of 2025 to 10.1% in the third quarter of 2026. But the business has only just begun to produce cash after stock pay. That is the main reason our valuation is well below what a multiple of adjusted EBITDA would suggest.
One item in the cash flow statement deserves attention. The company made $70.0 million of "loan advances" in the third quarter, recorded in other non-current assets. The 10-Q gives no borrower, term or purpose. It says only that the interest rate approximates a market rate. A loan of that size is about 30% of a year's free cash flow, with nothing disclosed about it. It sits outside the company's free-cash-flow figure, which deducts only capital expenditure and capitalised software from operating cash flow.
Testing the drivers against history#
Advertising cycles. Zeta's revenue is largely usage-based, so a slowdown in marketing spend should show up quickly. The last test was 2022–2023, when digital advertising budgets were cut across the industry. Zeta's revenue grew 29.0% in 2022 and 23.4% in 2023 (company figures of $458 million, $591 million and $729 million for 2021–2023, from the third-quarter 2026 supplemental; growth rates ours). The cycle slowed growth by about six points but did not stop it. One mild cycle is thin evidence. It does suggest that clients treat Zeta as part of their customer-retention spending, which is cut last.
Elections. Political candidates gave $44.4 million in fiscal 2024, 4.4% of revenue, and almost nothing in 2025. Growth excluding political and acquired revenue was 30% in 2024 and 27% in 2025 (fourth-quarter 2025 release). So the loss of election spending cost about three points of reported growth but did not change the underlying rate. In 2026 the company expects about $20 million, of which $7.4 million came in the third quarter.
Athena. Management names adoption of its AI agent as the driver of the 2026 acceleration. The CFO said on the November 9, 2026 call that Athena adopters were "over 70% of revenue" and that the top 25 users grew revenue "more than 70%". These figures cannot be tested yet: the company has not published how adopters spent before they adopted. The best check available is the organic rate. It moved from 28% to 30% over two quarters while adoption among super-scaled customers went from "just over 40%" to "over 50%". That is consistent with a modest effect, not a transforming one.
What management said that is new#
On the November 9, 2026 earnings call:
- 2027. The CFO said 2027 guidance, due in February, would start from organic growth of "20% to 21%" with "that same 2% to 5% cushion". He said the company has started each year at that level since its first guide in February 2022.
- Partnerships in 2027. The CEO said partnerships "have not been a material component of the growth" so far and "should add materially to growth next year".
- Palantir. The CEO repeated that the Palantir partnership "can add greater than $100 million in incremental annual revenue in the years to come". He gave no year. He said there were "eight yeses" from joint prospects and that two customers "are already actively operating". Neither the release nor the 10-Q quantifies any partnership revenue.
- Stock-based compensation. The CFO said it had fallen as a share of revenue "from 33% to 19% to 14% to 12% year-to-date", "and we expect that trend to continue". He gave no target.
- Tax. The release says guidance excludes a "one-time, non-cash tax benefit" from releasing part of the valuation allowance, which has "a reasonable possibility of occurring within the next 12 months". The company gave no amount. Such a release would lift GAAP net income for one period without any cash effect.
The tone has shifted from a marketing cloud to "intelligent AI infrastructure". The release now calls Zeta "the intelligent AI infrastructure company". The supplemental compares it with Palantir, Datadog and Snowflake. The financial reporting has not changed with the label. Revenue is still marketing revenue, from the same customers, priced the same way.
Management track record#
Guidance against results#
Beat = the actual figure is above the top of the range; met = inside the range; miss = below the bottom; a point guide is met within 1%. Scored is the last guidance given before the period ended (for a quarter, the range in the release that reported the quarter before it, or a later update). Revenue and adjusted EBITDA are scored for every quarter; free cash flow and GAAP EPS are scored for the full year when the company guided them.
Guidance, scored
22 of 22 scored guides above the top of the range (Q4 2024 to Q3 2026)
- Q4 2024:Beat
- FY 2024:Beat
- Q1 2025:Beat
- Q2 2025:Beat
- Q3 2025:Beat
- Q4 2025:Beat
- FY 2025:Beat
- Q1 2026:Beat
- Q2 2026:Beat
- Q3 2026:Beat
- Q4 2026:Due
- FY 2026:Due
All 22 scored rows are beats: eight quarters and two full years, from the fourth quarter of 2024 to the third quarter of 2026.
| Score | Period | Metric | Guidance | Actual | Note |
|---|---|---|---|---|---|
| Beat | Q4 2024 | Revenue, USD million | 293.0 to 297.0 | 314.7 | Given November 12, 2024, including LiveIntent from October 21, 2024 |
| Beat | Q4 2024 | Adjusted EBITDA, USD million | 64.9 to 66.9 | 70.4 | |
| Beat | FY 2024 | Revenue, USD million | 984.1 to 988.1 | 1,005.8 | First given February 27, 2024 at 870 to 880; raised three times |
| Beat | FY 2024 | Adjusted EBITDA, USD million | 187.5 to 189.5 | 193.0 | First given at 165 to 167 |
| Beat | FY 2024 | Free cash flow, USD million | 88 to 92 | 92.3 | First given at 75 to 85 |
| Beat | Q1 2025 | Revenue, USD million | 253 to 255 | 264.4 | |
| Beat | Q1 2025 | Adjusted EBITDA, USD million | 44.2 to 44.8 | 46.7 | |
| Beat | Q2 2025 | Revenue, USD million | 295 to 298 | 308.4 | |
| Beat | Q2 2025 | Adjusted EBITDA, USD million | 54.6 to 55.2 | 58.8 | |
| Beat | Q3 2025 | Revenue, USD million | 327 to 329 | 337.2 | |
| Beat | Q3 2025 | Adjusted EBITDA, USD million | 70.3 to 71.0 | 78.1 | |
| Beat | Q4 2025 | Revenue, USD million | 378.8 to 381.8 | 394.6 | Raised November 24, 2025 from 363 to 366 to add 15.8 of Marigold |
| Beat | Q4 2025 | Adjusted EBITDA, USD million | 90.7 to 91.5 | 95.1 | Raised November 24, 2025 from 89.7 to 90.5 |
| Beat | FY 2025 | Revenue, USD million | 1,289 to 1,292 | 1,304.7 | First given February 25, 2025 at 1,235 to 1,245 |
| Beat | FY 2025 | Adjusted EBITDA, USD million | 274.2 to 275.1 | 278.7 | First given at 255.5 to 257.5 |
| Beat | FY 2025 | Free cash flow, USD million | 156.9 to 157.9 | 164.7 | First given at 127.5 to 131.5 |
| Beat | Q1 2026 | Revenue, USD million | 369 to 371 | 396.3 | First given November 4, 2025 as 314.5 before Marigold; 362.0 on November 24, 2025 |
| Beat | Q1 2026 | Adjusted EBITDA, USD million | 61.2 to 61.8 | 66.1 | First given November 24, 2025 as 60.0 |
| Beat | Q2 2026 | Revenue, USD million | 419 to 422 | 442.8 | |
| Beat | Q2 2026 | Adjusted EBITDA, USD million | 86.2 to 86.9 | 91.7 | |
| Beat | Q3 2026 | Revenue, USD million | 469 to 472 | 492.7 | |
| Beat | Q3 2026 | Adjusted EBITDA, USD million | 115.0 to 116.0 | 121.0 | |
| Due | Q4 2026 | Revenue, USD million | 521 to 525 | Raised November 9, 2026 from a midpoint of 508 | |
| Due | Q4 2026 | Adjusted EBITDA, USD million | 135.2 to 136.3 | Raised November 9, 2026 from 131.9 | |
| Due | FY 2026 | Revenue, USD million | 1,853 to 1,857 | First given November 4, 2025 as 1,540 before Marigold; at least 1,730 with Marigold on November 24, 2025 | |
| Due | FY 2026 | Adjusted EBITDA, USD million | 414.1 to 415.2 | First given November 4, 2025 as 354.0 before Marigold; 385.4 on November 24, 2025 | |
| Due | FY 2026 | Free cash flow, USD million | 270.6 to 271.6 | First given November 4, 2025 as 209.0 before Marigold; 224.0 on November 24, 2025 | |
| Due | FY 2026 | GAAP diluted EPS, USD per share | 0.22 to 0.24 | Earlier 0.09 to 0.11 (August 4, 2026); excludes a possible tax benefit |
Source: earnings releases of November 12, 2024, February 25, 2025, May 1, 2025, August 5, 2025, November 4, 2025, February 24, 2026, April 30, 2026, August 4, 2026 and November 9, 2026; Form 8-K of November 24, 2025.
All 22 scored rows are beats: eight quarters and two full years, from the fourth quarter of 2024 to the third quarter of 2026. Quarterly revenue came in on average 4.4% above the top of the range, never less than 2.5% above it, and as much as 6.8% above it. Full-year beats are smaller, between 0.3% and 4.3% above the top of the last range, because by the fourth quarter little remains to be estimated.
The guidance is deliberately cushioned. The CFO said on the November 9, 2026 call that the outlook "includes a 2% to 5% cushion consistent with our typical conservatism". The steady size of the beats fits that statement. A beat of 4% is the plan working, not a surprise. The full-year 2026 path shows the effect over a year. Revenue excluding Marigold and political candidates was first guided at $1,525 million (November 24, 2025) and now stands at $1,635 million, 7.2% higher. Free-cash-flow guidance has risen 21.0%, from $224.0 million to $271.1 million at the midpoint.
The record is not one of luck. It is a planning convention, and its value lies in its stability. A beat smaller than about 2.5% above the top of the range would be the first sign that the cushion is being used up.
Promise register#
| Status | Promise, and what became of it | First made | Later restatements |
|---|---|---|---|
| Exceeded | Zeta 2025: revenue above $1 billion with an adjusted EBITDA margin of at least 20% by 20252025 revenue of $1,304.7 million at a 21.4% margin; revenue passed $1 billion a year early, in 2024 | February 23, 2022 release: "in excess of $1 billion in annual revenue with at least 20% Adjusted EBITDA margins by 2025" | |
| Raised | Zeta 2028: revenue of at least $2.1 billion, adjusted EBITDA of at least $525 million and free cash flow of at least $340 million in 2028Raised to at least $2.3 billion, $573 million and $371 million on February 24, 2026, mostly to include Marigold; 2026 guidance needs about 11% revenue growth a year to reach it | February 25, 2025 release: "Revenue of at least $2.1 billion; implied 20% organic CAGR" | February 24, 2026 release: at least $2.3 billion revenue, $573 million adjusted EBITDA, $371 million free cash flow |
| On track | Marigold to bring at least $190 million of revenue in 2026$151.7 million in the first nine months; the company guides $47.5 million for the fourth quarter, $199 million for the year | November 24, 2025 release: "at least $190 million from Marigold's enterprise software business" | November 9, 2026 supplemental: $199 million for 2026 |
| Open | Marigold to be accretive to adjusted EBITDA and free cash flow in its first yearNo profit figure for Marigold is disclosed; the first year ends in November 2026 | September 30, 2025 release announcing the Marigold agreement | |
| On track | Positive GAAP net income for the full year 2026$33.2 million in the first nine months; GAAP EPS guided at $0.22 to $0.24 | February 24, 2026 release: "Guiding to positive GAAP Net Income for the full year 2026" | Guided as $0.09 to $0.11 per share on August 4, 2026 and $0.22 to $0.24 on November 9, 2026 |
| Met | A sixth straight year of revenue growth above 20% in 20252025 revenue grew 29.7% | February 25, 2025 release | |
| On track | Buy back up to $200 million of shares by December 31, 2027$78.5 million remained at September 30, 2026 | The 2025 repurchase programme, as the Form 10-Q for the quarter ended September 30, 2026 records it | |
| Open | AthenaOS generally availableThe CEO said it is "in alpha now", with a beta for select customers in 2026 | November 9, 2026 earnings call: general availability in "the first quarter of 2027" | |
| Open | Palantir partnership to add more than $100 million of annual revenueNo revenue disclosed; the CEO cited "eight yeses" and two customers live | November 9, 2026 earnings call: "greater than $100 million in incremental annual revenue in the years to come" |
What the ten status words mean
- Met
- Done as promised, by the date given.
- Exceeded
- Done, and more than promised.
- On track
- Not due yet; the latest figures point to it.
- Open
- Not due yet, and the figures so far cannot settle it.
- Slipped
- Not done by the date given, and still pursued.
- Lowered
- Cut openly before its date; the new target is in "What became of it".
- Raised
- Lifted openly before its date; the new target is in "What became of it".
- Quietly reframed
- Changed without being called a change.
- Dropped
- No longer pursued or mentioned.
- Missed
- Due, and not delivered.
Source: earnings releases of February 23, 2022, February 25, 2025, February 24, 2026, August 4, 2026 and November 9, 2026; Forms 8-K of September 30, 2025 and November 24, 2025; Form 10-Q for the quarter ended September 30, 2026; third-quarter 2026 earnings supplemental; Zeta third-quarter 2026 earnings call, November 9, 2026.
The register leaves out smaller product launches and the targets for individual metrics in the 2028 model, which are covered in the guidance discussion above. Zeta 2028 is marked Raised because the target was lifted before its date. The new revenue target of $2.3 billion needs about 11% growth a year from the 2026 guidance. The organic rate has been running at 28–30%. By February 2027, when the company next guides, the target will look conservative unless growth collapses.
When targets moved#
No target in the register has slipped or been missed. The raise of Zeta 2028 in February 2026 came from an acquisition, Marigold, not from faster organic growth, and the company said so. The original organic aim of a 20% compound rate is still the company's planning base. The CFO's description of the 2027 starting point ("20% to 21%") repeats it.
The one soft spot is Marigold itself. Its revenue fell from $55.6 million in the first quarter of 2026 to $48.1 million in the second, and the company has not explained the fall. The $190 million promise for 2026 is still on track only because the fourth-quarter guide of $47.5 million needs no recovery. Whether Marigold was accretive in its first year, as the September 2025 release promised, cannot be checked: the company does not report its profit.
Capital allocation#
| Period | Shares repurchased | Cost, USD million | Average price, USD per share |
|---|---|---|---|
| FY 2025 | 7,899,208 | 120.1 | 15.20 |
| Q1 2026 | 1,549,368 | 25.7 | 16.60 |
| Q2 2026 | 1,647,622 | 29.9 | 18.15 |
| Q3 2026 | 1,280,261 | 29.9 | 23.34 |
The buybacks were well priced against later prices. The average of $19.10 per share in the first nine months of 2026 is 41% below the $32.42 close of November 9, 2026. But they did not stop the share count from rising:
- Class A and B shares outstanding rose 3.2% in 2025 and another 2.9% in the first nine months of 2026, to 251.9 million (our calculation from the company's share counts).
- The $85.5 million spent on buybacks in the nine months was 55% of that period's stock-based compensation.
- Marigold was paid partly in shares: 5,329,070 at closing and 3,905,904 to settle the seller notes. Together that is 3.8% of the shares outstanding at the end of 2025, which we calculate.
The October 2026 acquisitions of Senso (about $27.5 million) and Digital Audience (about $35 million) add a further 694,582 shares, plus earn-outs payable in shares. The company also registered up to $26.25 million of shares to pay a service provider (prospectus supplements of September 11 and October 16, 2026). The 10-Q records an obligation to issue $40.0 million of shares under a "long-term strategic partnership agreement".
Debt is modest. A $250 million term loan was drawn under a new $1.0 billion facility on July 24, 2026, against $340.1 million of cash at September 30, 2026. Management has said it prefers "small to mid-sized" acquisitions (CEO, August 4, 2026 call).
Insiders#
The Forms 4 the company's officers filed over the past year show no open-market purchases. They also show no open-market sales by the CEO, the CFO or the President. Those officers instead transferred shares as gifts to trusts managed by independent trustees.
On August 13, 2026, Botticelli SPV LLC entered into a variable prepaid forward contract. Botticelli is owned by a trust of which the CEO and his children are the beneficiaries, and the CEO has sole voting power over its shares. It received $22.7 million upfront and pledged 1,000,000 Class B shares. The contract settles after August 13, 2029, with a floor price of $26.11 and a cap of $47.29 (Form 4 filed August 14, 2026). Botticelli keeps the voting rights on the shares during the contract. The contract locks in a price range for part of the holding and turns it into cash now, which is economically close to a partial sale.
Read together, the record shows executives reducing their exposure through trusts and forward contracts, not through direct sales. It is disclosed in the filings. It also means insiders have not been adding to their holdings at any price.
How far management's word can be trusted#
On the figures it guides (revenue, adjusted EBITDA, free cash flow), the record is as clean as a record gets: 22 of 22 scored rows above the top of the range. The cushion is stated openly. On long-term targets, Zeta 2025 was met early, and Zeta 2028 is within reach on 2026 numbers.
The record is weaker where management does not guide or report:
- Marigold's profit.
- The terms of a $70.0 million loan.
- Any revenue from Athena, Palantir or OpenAI.
- How the description of the consumer data set came to change.
Our judgement: trust the guided numbers as floors, and treat the narrative claims (AI infrastructure, the Palantir figure) as unverified until revenue lines appear in the filings.