DMRC · Research
Digimarc Lost 27 Percent of Recurring Revenue and Its 10-Q Doubts It Can Fund a Year
Digimarc Corporation has strong watermark technology and a 30-year central bank customer, but owner cash is deeply negative, an activist holds 18.5% and new shares keep coming.
- Owner cash after capex, patents and all stock pay
- Minus 85.9% of revenue
- Annual recurring revenue
- Down 27%
- Share count growth since 2021
- About 60%
1. Our call: Avoid, because the company's own 10-Q now doubts it can fund a year
Our call: Avoid. Digimarc Corporation sells digital watermarking software that hides machine-readable identifiers in packaging, labels, images and banknotes, used for product authentication, retail loss prevention and recycling sortation [1].
Its oldest customer is a consortium of the world's central banks, which has used the technology against currency counterfeiting for almost 30 years [1].
This is a small pick. Revenue was $33.9M in 2025, down 12%, and $15.0M in the first half of 2026, down 14% [1][2].
We chose it because it is confirmed software, it posts results on Nov 12, outside our seven-day window, and it is a U.S. company with a U.S. team, but readers should weigh its size [12][1].
The decisive fact is in the second quarter 10-Q. Management states that cash and marketable securities of $8.8M at Jun 30, 2026 will not fund operations for the next twelve months, and that substantial doubt exists about the company's ability to continue as a going concern [2].
Since then the company raised about $18.2M gross in a registered direct offering on Sep 30, 2026
2. A watermark vendor whose gate never charged a toll
| Metric | Value |
|---|---|
| Recurring share of revenue | Subscription was 59% of 2025 revenue and 54% of first half 2026 revenue, our calculation [1][2] |
| Operating margin | Minus 97.8% in 2025, operating loss of $33.2M on $33.9M, our calculation [6] |
| Stock pay share of revenue | 35.3% in 2025 ($12.0M), 66.4% in the first half of 2026, our calculation [6][2] |
| API quality | 2 of 5, our scoring: the platform offers detection SDKs and integrations, but adoption depends on printers and retailers [1] |
| Toll-gate score | 0 of 4 pass, 2 partial, 2 fail (section 2b) |
2b. Digimarc can mark the object, but others own the checkout lane
A gate is valuable when its owner holds the record others depend on and prices what passes through. Digimarc's identifiers sit on banknotes and packages, but the scanners, the retail lanes and the recycling lines belong to customers, and most commercial pilots have not become broad paid deployments [1][2].
| Test | Result | Evidence |
|---|---|---|
| Owns the record | Partial | The watermark and its resolver are Digimarc's, and the central bank work spans almost 30 years, but brands and retailers own the product data the marks point to [1]. |
| Owns the connection | Partial | Detection is built into some scanners and devices, but annual recurring revenue fell to $11.6M from $15.9M after a large contract expired and another stepped down [3]. |
| Prices close to outcomes | Fail | Subscriptions are fixed fees over one to three years, not tied to recovered losses or verified items [1]. |
| R&D on the gate vs features | Fail | Research spending fell to $7.4M in the first half of 2026 from $12.2M a year earlier as the company cut costs, not as the gate gained users [2]. |
3. Revenue peaked in 2024 and has fallen since
Subscription revenue carried the decline. It was $3.7M in the second quarter of 2026 against $4.6M a year earlier, while service revenue rose to $3.6M from $3.4M [2].
Service is mainly software development for the central banks, so the mix has moved back toward the oldest customer [1][2].
Annual recurring revenue tells the same story. It was $11.6M at Jun 30, 2026 against $15.9M a year earlier [3].
The release attributes $3.1M of the drop to a commercial contract that expired in October 2025 and $2.6M to a step-down in June 2026 [3].
Government revenue, chiefly the central banks, is international revenue in the filings and the most stable line Digimarc has [1][2]. A business that cannot grow its commercial base leans harder on a single consortium contract each year
4. There is no guide, and the next report comes Nov 12
Digimarc does not give revenue guidance in its results release [3]. The company has said its priorities are reducing cash use and focusing on the products with the shortest path to revenue [3].
Free cash flow usage, a company measure, was $1.0M in the second quarter, much smaller than the reported net loss [3].
That gap is mostly stock pay and noncash charges, and it is real cash relief [3][2]. But it came alongside a 27% drop in recurring revenue, which suggests cost cuts are buying time rather than building a business [3].
Digimarc posts third quarter 2026 results on Nov 12, 2026 [12].
5. Margins show a software core inside a loss-making company
Gross margin was 58% in the second quarter of 2026, with subscription gross margin of 89% [3]. The subscription product is economically sound at the unit level [3].
Operating costs are the problem. Operating expenses were $16.7M in the second quarter against $13.1M a year earlier, and general and administrative expense rose to $10.3M from $5.1M on chief executive transition charges [2][3].
The net loss was $12.1M on revenue of $7.4M [3].
Operating losses were $39.9M in 2021, $61.8M in 2022, $48.2M in 2023, $41.3M in 2024 and $33.2M in 2025 [6]. Digimarc has not had a profitable year in the period we reviewed [6].
6. Owner cash has been deeply negative every year
| Period | Operating cash | Capex and patents | Stock pay | Owner cash | Share of revenue |
|---|---|---|---|---|---|
| 2023 | Minus $22.0M | $0.7M | $11.2M | Minus $33.9M | Minus 97.3% [6] |
| 2024 | Minus $26.6M | $0.6M | $10.0M | Minus $37.2M | Minus 96.9% [6] |
| 2025 | Minus $11.8M | $1.2M | $12.0M | Minus $25.0M | Minus 73.6% [6] |
| First half 2025 | Minus $10.2M | $0.5M | $5.0M | Minus $15.7M | Minus 90.1% [6] |
| First half 2026 | Minus $2.6M | $0.3M | $9.9M | Minus $12.9M | Minus 85.9% [2] |
Owner cash is our calculation: operating cash flow less purchases of property and equipment and capitalized patent costs, less all stock pay expense [6][2]. Stock pay in the first half of 2026 includes about $5.4M of accelerated expense on the former chief executive's awards [2][8].
Even excluding that charge, owner cash stays far below zero [2][6]. Operating cash use did shrink sharply in 2026, which is the one real improvement in the record [6][2].
7. Capital: dilution is now the operating plan
Shares outstanding were 16,846,589 in February 2021 and 22,475,329 on Aug 7, 2026 [6][2]. Adding the 4.5M shares sold on Sep 30, 2026, the count is about 27.0M, an increase of about 60% since 2021, our calculation [9][10].
On Jun 8, 2026 Digimarc signed an at-the-market sales agreement with Needham & Company for up to $17.5M of stock, paying a 3% commission [7]. The Sep 30 registered direct offering raised about $18.2M gross and gave buyers participation rights in future offerings for up to three years [9].
The company also repurchased $1.5M of stock in the first half of 2026, while it was preparing to raise equity [6][2]. The binding constraint is cash runway, as the 10-Q itself states [2].
8. The main risks are runway, a shrinking base and governance churn
Runway. The 10-Q going concern statement predates the Sep 30 raise, and the raise roughly triples the Jun 30 cash balance, but quarterly operating cash use and planned spending decide how long it lasts [2][9].
Shrinking base. Annual recurring revenue fell 27% in a year, two large commercial contracts drove most of the drop, and the company has not named replacements [3].
Key-person and governance churn. The founder-era chief executive Riley McCormack left in July 2026, and the new chief executive started on Jul 6, 2026 [7][13].
The largest holder now has a stated agenda to replace legacy directors [10].
Dilution. Stock pay was 35.3% of 2025 revenue, the new chief executive received 1,060,000 units, and equity raises are continuing [6][7][9].
9. Management wants to be the trust layer for physical and digital goods
The 10-K frames Digimarc as a digital trust company for anti-counterfeiting, retail loss prevention and recycling, built on watermarking it has developed for decades [1]. The April 2025 Holy Grail 2.0 trial in Germany validated watermark-based recycling sortation at industrial scale [1].
If regulators or large retailers mandate digital marks, Digimarc owns a technology and patent base few rivals match [1]. If not, it remains a vendor of pilots, and the central bank contract funds the core engineering [1][2].
10. Superintelligence raises the value of authenticity and the threat to it
Synthetic images and counterfeits generated by superintelligence tools make provenance more valuable, and Digimarc markets its marks for digital content authenticity [1]. That is the bull case for the gate in section 2b [1].
The same tools can strip or forge visible signals, and large platforms are building their own content credentials [1]. A small vendor with 110 employees competes against platforms that can embed provenance by default [1].
11. An activist at 18.5%, a departed founder-investor and a new chief executive
Vote control. Digimarc has one class of voting common stock and no controlling holder [4].
The ten shares of nonvoting Series A preferred carry no economic rights and are redeemable at the board's option [1]. On May 15, 2026 shareholders became holders of Digimarc Parent, Inc., a new holding company, and the operating business became DMRC LLC, so that executive awards can be issued as units in the subsidiary [4][2].
The proxy gave four reasons for the reorganization: cash savings, less dilution, retention of key talent and tying awards to shareholder value [4]. It passed with 98.8% of votes for or against, our calculation [5].
Activists. Ocho Investments LLC, a Delaware company managed by Andris Upitis of Austin, Texas, reported 5,000,000 shares, 18.5% of 27,048,451, on Sep 30, 2026 [10].
It bought 3,675,000 of the 4.5M offering shares for $14.8M, after holding 1,325,000 in July
12. The questions we would ask management, and the calls we are tracking
Question one: how many quarters of operating cash use does the Sep 30 raise fund, and does the going concern language come out of the third quarter 10-Q? [2][9]
Question two: what replaces the $5.7M of recurring revenue lost to the October 2025 expiry and the June 2026 step-down? [3]
Question three: what share of revenue comes from the central banks, and when does that agreement renew? [1]
Question four: has the board discussed board seats or a sale with Ocho Investments LLC? [10]
Question five: why grant over one million units to a new chief executive at a company that doubts its own runway? That answer sets our conviction.
Bottom line: good technology, a shrinking base and a runway funded by dilution
Digimarc has 89% subscription gross margin, decades of watermark patents and a central bank customer of almost 30 years [3][1]. But owner cash was minus 85.9% of revenue in the first half, recurring revenue fell 27%, the 10-Q raised substantial doubt about the going concern, and the share count is up about 60% since 2021 [2][3][9].
We would move to Hold off if recurring revenue grows for two straight quarters and the going concern language is removed, and stay at Avoid on any new raise in 2027. Our call: Avoid, conviction 4 of 5.
Stock Profiles & Comparable Companies
DMRC Stock Profile →More Published Research
View All →i3 Verticals Borrowed to Retire a Quarter of Its Class a Stock While Growth Slowed to 3 Percent
A founder-led public sector software company that drained its cash for buybacks, pays insiders under a tax agreement, and lets the founder's vote rise as it shrinks.
Octave Wrote Off Over 2 Billion Dollars in Its First Public Quarter and One Swedish Family Holds 43 Percent of the Votes
Hexagon's industrial software spin earns owner cash of about a fifth of revenue, but the top line is flat, controls are not effective and a 10-vote share class keeps control in Stockholm.