Veritone, Inc. (VERI) Earnings Call Transcript & Summary
March 26, 2026
What were the key takeaways from Veritone, Inc.'s March 26, 2026 earnings call?
In the preliminary unaudited results for Q4 2025, Veritone, Inc. reported a revenue range of $18.1 million to $30 million, heavily influenced by a strategic $12.9 million transaction involving their aiWARE stack. The company is optimistic about future revenue, particularly from a new partnership with Oracle and access to over 50 million hours of monetizable data, projecting fiscal 2026 revenues between $130 million and $145 million, representing a 47% year-over-year increase. Management highlighted a strong growth outlook for the public sector, targeting 60% to 70% growth, while also indicating a potential net loss improvement for the year.
What topics did Veritone, Inc. cover?
- Strategic Data Transaction: Veritone completed a significant transaction valued at $12.9 million, which includes access to valuable data sets for future monetization. Management stated, "We now have preferred access to potential VDR customers, who control more than 50 million hours of monetizable data sets today."
- Partnership with Oracle: A multiyear strategic partnership with Oracle was announced to enhance the deployment of Veritone's aiWARE platform. This partnership is expected to provide substantial financial incentives and improve operational efficiency, as noted by management: "This partnership ensures that Veritone's customer base can harness the transformative potential of AI with superior price performance, security and data sovereignty."
- Public Sector Growth: Veritone's public sector division is projected to grow between 60% to 70% year-over-year, driven by existing federal contracts and new international deals. Management emphasized, "The TAM for digital evidence management solutions today exceeds $10 billion and is growing at double-digit rates."
- Revenue Guidance for FY 2026: Management provided guidance for fiscal 2026, expecting revenues between $130 million and $145 million, which represents a 47% increase from the low end of the 2025 range. They stated, "We expect to see a more pronounced revenue ramp in VDR starting in Q2 and throughout the second half of 2026."
- Operational Efficiency Improvements: Veritone has made significant strides in operational efficiency, retiring 100% of senior secured term debt and reducing annual debt carrying costs by 90%. CFO Mike Zemetra noted, "We closed fiscal year 2025 with unencumbered cash and cash equivalents of $27.7 million."
What were Veritone, Inc.'s March 26, 2026 results?
- Revenue Range Q4 2025: $18.1M - $30M (vs $25M est, driven by a strategic transaction)
- Projected Revenue FY 2026: $130M - $145M (represents a 47% increase year-over-year)
- Public Sector Growth Rate: 60% - 70% (year-over-year growth target)
- Net Loss Guidance FY 2026: $13.5M - $22.5M (54% improvement year-over-year at the low end)
- Debt Reduction: 90% (reduction in annual debt carrying costs from $14M to $800K)
- VDR Pipeline: $50M+ (near-term sales pipeline in VDR)
Veritone's strategic initiatives, including the Oracle partnership and access to extensive data sets, position the company for significant growth in 2026. However, the challenges in data sourcing and the complexities of government contracts present risks. Investors should monitor the execution of these strategies and the ramp-up of revenue from the VDR segment.
Earnings Call Speaker Segments
Thanks, and good afternoon, everyone. Veritone put out a press release once the market shut down today, and in it we announced preliminary unaudited financial results covering both the fourth quarter and the entire year that wrapped up on December 31, 2025. You can find that release in the investor relations part of Veritone's site. With us on today's call are Ryan Steelberg, our President and Chief Executive Officer, and Mike Zemetra, our Chief Financial Officer, and they'll be giving prepared remarks. I want to point out that some of what gets discussed today falls under forward-looking statements. That covers, among other things, statements regarding our business strategy along with future financial and operating performance. Risks, uncertainties, and assumptions apply to these forward-looking statements, and they could cause actual results to come in materially different from what's stated. Some of those risks and assumptions get covered in Veritone's SEC filings, such as its annual report on Form 10-K. These forward-looking statements rest on assumptions as of today, March 26, 2026, and unless a subsequent press release tied to the company's audited Q4 and full year 2025 financial results says otherwise, Veritone has no obligation to revise or update them. On this call, the actual and forecasted financial measures we discuss will include non-GAAP measures. You'll find reconciliations of those measures against the corresponding GAAP measures in the press release we issued today. Last thing — a reminder that today's call is being recorded and can be replayed through a link on the investor relations section of Veritone's website at www.veritone.com. At this point, I'd like to hand the call over to our President and Chief Executive Officer, Ryan Steelberg.
Good afternoon, everyone. I'm glad to share where we stand on our preliminary Q4 financials, talk through the strategic deal we announced today with Oracle, and offer a deeper strategic look at some of our newer, exciting business opportunities, including the meaningful headway we've made with certain hyperscalers. Since these results are preliminary, there won't be a Q&A session after today's prepared remarks. The preliminary Q4 figures we furnished today span a fairly wide revenue range, from $18.1 million to $30 million. A single transaction, which we signed and completed in Q4 2025, accounts for nearly all of that spread. This deal was a complex arrangement among several parties with no cash changing hands, and it centered on an on-premises transaction for our aiWARE stack and application software, valued at $12.9 million. In exchange, we received specific intangible rights along with direct and preferred entry to a large base of customers who control a range of digital datasets, which we'll use going forward for VDR and AI model training, structured under a fixed 50% revenue share — a notable improvement over the margin we presently make on VDR. Although the software's contracted price came to $12.9 million, from a pure GAAP accounting standpoint, determining the proper fair value of the sale is quite difficult, since realizing the data rights lies in the future, given how relatively new — though rapidly expanding — the markets for data sales and VDR are. Consequently, the standalone selling price of the software might be discounted considerably from a revenue standpoint. Worth noting: that $12.9 million figure was part of our earlier Q4 revenue guidance range, yet it may land at the lower end of this range. It's very much binary. Even so, we're putting in significant effort, using teams inside the company as well as outside partners, to ensure the worth of this is properly captured when we report Q4 2025. Our expectation is that this effort will be finished before we submit our 10-K. No matter what the ultimate GAAP treatment turns out to be, I'd like to explain why this transaction is so important to Veritone from a strategic standpoint, and why we believe that in time we will turn this active deal, which carries a $12.9 million face value, into something considerably larger. Right now, Veritone has both exclusive and non-exclusive rights and relationships that let us make money from some of the most valuable and iconic datasets around — in sports, entertainment, public safety, and news. Those come from content and IP owners like the NCAA, CBS News, and other franchise customers. That being said, demand for VDR-specific content has gone up dramatically, and what content is needed has shifted across a very wide spectrum. As it stands, under the contracts Veritone has today, we don't have universal access to an unlimited supply of data suppliers. To give you an example, in fiscal 2025 we had to turn down more than $10 million in bona fide data orders, since we couldn't source the volume required, or the specific type of content or data, within the time frames that our hyperscaler and model developer partners were asking for. This strategic agreement is one of the ways we're moving forward, and it fits together with the business development efforts we already run day to day. As a result of the deal, we've got preferred access right now to potential VDR customers who collectively control more than 50 million hours of datasets that can be monetized. To put that in perspective, the entire marquee catalog held in the NCAA video library adds up to less than 1% of that figure. As one example of the new customers coming out of this transaction — and we signed an agreement with them not long ago — a large fast food franchisee is providing Veritone with access to its collection of multi-camera video surveillance footage, which is exactly the kind of data hyperscalers want badly for frontier and world model development. Based on what we're seeing, our projections take a cautious view: the full set of datasets tied to this deal, when offered to prospective VDR clients, has the potential to generate over $100 million in VDR revenue within the coming three years. Beyond that, we earn a set margin on this data, and it's substantially stronger than the VDR margin we've seen historically. Even though accounting rules may mean that revenue gets recognized gradually down the line, what we have right now is access to a large pool of valuable datasets we can turn into money through our VDR platform, and we didn't have to spend any cash upfront to get them. Monetization of this dataset should get underway as soon as Q2 2026, in our view. I realize I've devoted considerable time to this transaction, but it's important to us that our investors fully grasp both the situation and the major opportunity it presents. Now, I couldn't be more thrilled about what comes next. I want to introduce and talk through the revolutionary agreement we've just unveiled with Oracle — one that took nearly twelve months to finalize. This partnership with Oracle is a multi-year strategic arrangement aimed at speeding up how Veritone's aiWARE platform, application, and data services get deployed through Oracle Cloud Infrastructure, or OCI, and it comes with embedded, sizable cash incentives that let Veritone expand its cloud infrastructure while achieving more cost-effective compute and storage. This deal makes Oracle Cloud Infrastructure the foundation for Veritone's upcoming AI offerings across commercial, public sector, and fast-growing Veritone Data Refinery markets. With demand for AI infrastructure that scales and stays secure hitting levels never seen before, Veritone plans to tap OCI's high-performance AI superclusters to drive its aiWARE platform and data solutions. Through this partnership, Veritone's customers gain access to AI's transformative power, backed by the strong price performance, security, and data sovereignty that come with Oracle's distributed cloud. For Veritone, teaming up with Oracle marks a major step forward in its effort to help enterprises uncover the truth hidden in their data. The financial commitments Oracle is making as part of this partnership serve as proof that we lead when it comes to handling unstructured data, and that we are dedicated to delivering among the strongest AI offerings available anywhere. By moving mission-critical workloads onto OCI, we unlock new levels of scalability for Veritone Data Refinery, while giving both our government and business customers the speed and protection required to stay out in front in an AI-powered world. You can read more in the joint press release available on our investor site. Let's now move on to discuss our core operating business. Across the last 2 years, we've followed a disciplined path toward realigning Veritone behind one clear vision, restoring aiWARE as the indispensable orchestration and intelligence layer for enterprise and public sector AI along with unstructured data. Today, I'm pleased to share that Veritone's transformation is finished. We've rebuilt this organization into a focused, platform-driven company, and we're now executing from a place of operational, financial, and technical strength. During 2024, we began repositioning the company. During 2025, we validated that model, and as of today we're no longer managing our way through a transition. We've exited non-core assets, we've streamlined our operating structure, and we've aligned our resources behind scalable, platform-driven revenue streams. Right now, we're putting money to work, expanding what we've built, and producing the kind of operating leverage you'd expect from a company leading its market. Nothing demonstrates how far we've come better than the Veritone Data Refinery — VDR for short — and what matters even more is the group of hyperscale, best-in-class clients we're serving today.
What started out as a distinctive capability of our platform, one we built organically, has grown into a scaled engine for token production that today powers the leading hyperscalers and foundational model developers around the globe. Converting proprietary unstructured datasets into recurring revenue was our strategy, and the highest levels have confirmed it works; we're now actively engaged, under contract, and bringing in revenue with every major player, such as Nvidia, Amazon, Google, and Meta, among others. Given that we've expanded our access to and supply of rich data and have contracts in place with every major hyperscaler, we see an obvious chance to deepen and broaden our engagement with each of these partners individually. With hyperscalers projected to spend roughly $700 billion combined on CapEx in 2026, per S&P, Veritone is positioned to claim a growing portion of that investment as a top provider of clean, model-ready training data. When it comes to supply, we're effectively allowing enterprises to treat their unstructured audio, video, images, and text as a renewable, monetizable asset class that meets this AI demand. As we've said before, staying highly focused and cost-efficient on data acquisition is essential. On the VDR supply side, beyond the datasets that are now available to us thanks to the strategic transaction I described earlier, we keep signing and growing our portfolio of available datasets via other channels, and we keep enhancing our ability to refine, process, and prepare those datasets for licensing, distribution, and sales. In the past, sales and representation depended heavily on business development and manual work, which, though effective, held back our capacity to accelerate substantially in acquiring and distributing our dataset offerings and portfolio. That's no longer the case. I'm glad to build on the announcement from a couple weeks ago regarding the launch of the Veritone Data Marketplace built on aiWARE. Veritone has formally advanced to the next stage in the AI supply chain's evolution. Embracing the data economy, we recently announced the launch of the Veritone Data Marketplace, a platform built to enhance and streamline the data ecosystem while preserving control, transparency, and quality. This marketplace is the indispensable counterpart to our Veritone Data Refinery. While VDR functions as the tokenization engine that renders unstructured data AI-ready, VDM, the Veritone Data Marketplace, acts as the trusted and accredited distribution hub, making sure those assets are monetizable, transactional, and governed. We're especially enthusiastic about the value this delivers to both sides of the data licensing market. Rights holders who own valuable archives and datasets now have a straightforward route to monetization with full asset control. At the same time, AI developers can tap a deep catalog of proprietary data covering industries from entertainment to human behavior to robotics. These datasets may even be assessed for metadata density and model fit prior to acquisition, guaranteeing they're optimized for foundational architectures. The Veritone Data Marketplace used to be open to just a small set of customers; now content owners and AI developers can use it too, speeding up responsible AI work. We take pride in being out front when it comes to ethically sourcing top-quality data and offering it via a governed, transparent marketplace that serves what the hyperscaler community needs. Our public sector division is beginning to fire on every cylinder, posting robust 68% growth compared with a year earlier; three clear strategic pillars underpin this achievement. The first is high-impact applications. Our Veritone iDEMS suite is transforming productivity, powering mission-critical workflows that would be impossible without Veritone. That has sharply broadened our addressable market spanning SLED, higher education, FedCiv, and international agencies. The second is unmatched deployment flexibility. aiWARE and iDEMS are built for the most sensitive environments; we satisfy the toughest security and sovereignty requirements, whether running in government clouds or in fully network-isolated air-gapped environments. Thanks to Oracle, we'll elevate that to an even greater level of performance and security, at home and around the world. The third is a genuinely open architecture. In contrast to numerous rivals, aiWARE is an entirely open platform. Our capacity to ingest data at scale and link with any application or dataset free of vendor lock-in makes us the go-to infrastructure for federal AI monetization and the Department of Defense's AI First strategy. This groundwork has produced deep integration inside the Air Force OSI and the JPATS Trust Modernization Program; with our pipeline at record levels, our Q4 wins, among them a major U.S. university, a top-five sheriff department, and another major state highway patrol, genuinely confirm that we are a trusted AI partner for the public sector. On the commercial enterprise side, we have effectively operationalized our data-to-AI flywheel. By linking proprietary supply with surging AI demand, we created a scalable architecture in which volumes drive value. Additional data draws additional buyers, which directly powers our margin expansion and relicensing efficiency. The results say it all. During Q4, our licensing team closed 224 orders, up almost 10% versus the prior year. Our reach is genuinely global, supplying essential media assets to top-tier firms such as Google and Goldman Sachs, major studios such as NBCUniversal, and premier sport entities including ESPN, the NFL, and Tom Brady's Religion of Sports; this momentum directly feeds our high-margin software business. Software deal volume rose 14% year-over-year to 33 deals in Q4; we are witnessing a potent mix of retention and expansion, renewing core partners such as Sony Pictures and Summit Media, while at the same time landing new accounts such as Snap and deepening our relationships with prestigious events such as the London Marathon and Augusta National Golf Club. I'm extremely proud of our hire division, now rebranded as Broadbean by Veritone, which delivered yet another strong year. Despite macro hiring headwinds, Broadbean maintained and contributed robust profit margins critical to Veritone's overall growth and success. To grasp the scale that our Broadbean division manages, in 2025, Broadbean distributed over 7.6 million unique jobs, powering more than 40 million unique job ads. We ended up generating a remarkable 132 million engagements in application, along with clicks that go straight into the ATS and recruitment systems our customers use. On a monthly basis, roughly 30,000 distinct HR professionals depend on our platform for handling their talent acquisition. Among other standout items, our global media services division had a record-setting year, posting double-digit net revenue growth compared with the prior year; in fact, Q4 stood as our best ever, the first occasion we went past the $10 million mark for media under management. That momentum has come right along with us into 2026. I'm delighted to share news of a significant SaaS victory with the U.K.'s Department for Work and Pensions. Through this partnership, Veritone becomes a foundational piece of U.K. government recruitment, since our software will now drive job advertising for the Ministry of Justice, Defra, and the Home Office. On top of that, our inaugural year in the Workday Platinum Partner Program proved a success, landing 59 new deals — a 30% gain versus the year before. With a growing pipeline of Fortune and Forbes Global 500 brands, we're only at the beginning. Going forward, we're getting ready to introduce our next-generation job management modules along with our pioneering agentic AI Broadbean framework. This is more than simply an upgrade. We see it as, once again, a productivity revolution for the tens of thousands of people who use Broadbean each month. Here's to 2026.
Heading into this new year, our priority is straightforward: scaling with discipline. Our attention is on turning more than $50 million in VDR pipeline into recurring revenue we can recognize, while also growing our deployments in the public sector. We'll keep directing capital toward expanding the platform, and we'll keep assessing chosen strategic chances and partnerships that reinforce our moat in data and orchestration. The transition period is behind Veritone now. With our platform, our partners, and our financial base, we're positioned to lead the data economy powered by AI. That's all from me. At this point, I'd like to hand things to Mike Zemetra, our Chief Financial Officer. Mike.
Thanks, Ryan. Because these financial results are still preliminary, I'll limit my remarks today to our fiscal 2026 guidance and a handful of balance sheet updates. Across our software, products, and services, both the revenue pipeline and the long-term outlook remain at record levels. To be specific, demand stays strong in commercial VDR and in the public sector. In 2026, hyperscalers such as Google, Amazon, Meta, Nvidia, and others have each projected spending in the hundreds of billions during fiscal 2026 to fuel their AI efforts, which includes additional investment in their large language models. Fortune Business Insights values the worldwide AI training dataset market at roughly $3.6 billion in 2025, and expects it to rise from $4.4 billion in 2026 to $23.2 billion by 2034, representing a 23% CAGR. On the model training side, we think we're positioned well to capture this possible revenue opportunity, since more mature models are now putting substantial money into rich video data, an area where we believe Veritone holds a distinct competitive edge. As of today, our near-term sales pipeline for VDR by itself exceeds $50 million and keeps expanding. A key lesson from 2025 was that we needed to widen the range of dataset availability for our VDR customers and boost the speed at which those datasets get deployed. As Ryan noted, in fiscal 2025 we couldn't capture millions of dollars in possible VDR revenue purely because we couldn't readily obtain the content our VDR customers asked for quickly enough. To fix that in 2026, we'll concentrate on the most efficient, cost-effective approaches to expanding data supply, and we'll invest further in engineering and product for VDR, including Veritone Marketplace, with the goal of deepening our competitive moat through exclusive access to thousands of additional data providers. As Ryan explained, our Q4 strategic dataset transaction gives us access to customers controlling over 50 million hours of valuable video datasets, among them some of the biggest retail, travel, entertainment, and fast food providers globally; we believe these near-term strategic choices will let us keep growing VDR revenue in fiscal 2026 and beyond at or above the current 23% projected CAGR for large language model spending through fiscal 2034. In the public sector, the TAM for digital evidence management solutions now surpasses $10 billion and is expanding at double-digit rates. For fiscal 2026, we're aiming for our public sector to grow 60%-70% year over year. That growth is expected to come from expanded offerings under existing federal contracts, among them those with the DLA and OSI, plus new international deals throughout Western Europe. Altogether, our backlog and sales pipeline across our core AI platform today is greater than $200 million. Veritone stays uniquely positioned to seize even more opportunity in the data as a currency market. We anticipate that our pipeline, along with our ability to generate revenue from our extensive collection of tokenized audio and video assets, will expand even more. Regarding operating expenses, we anticipate sales, marketing, and G&A costs to remain relatively unchanged in fiscal 2026 versus 2025, and we expect spending in these categories as a proportion of total revenue to improve compared with the prior year. We project that research and development expenses will be somewhat greater than the previous year on an absolute dollar basis, beginning in Q1 and continuing through fiscal 2026, as we keep investing in our VDR and public sector revenue efforts, such as the Veritone Marketplace and planned new software product features and enhancements in 2026. Keep in mind that, consistent with 2025, we anticipate revenue will rise sequentially from quarter to quarter in 2026, with Q1 2026 revenue coming in close to Q1 2025; this stems partly from the public sector, where we foresee a larger revenue ramp beginning late in the first half of 2026 from our existing bigger federal deals, along with the timing of certain international contracts we expect to announce in the coming year; furthermore, based on discussions and the timing of certain VDR deals, including the signing of several large hyperscalers in late Q1 2026, we anticipate a more pronounced revenue ramp in VDR beginning in Q2 and continuing throughout the second half of 2026. The primary risks to our revenue projections involve the consumption-based nature of VDR, together with the timing of government contracts and decision-making. Additionally, visibility into our VDR pipeline is usually two to three months ahead of delivery, and decisions about the nature and volume of content may shift depending on the customer's needs and the anticipated effect on its training models. Because of these factors, along with the complexity of government decision-making, particularly during wartime, we will only provide financial guidance for fiscal year 2026, which we intend to update on our next earnings call. To be more specific, in fiscal 2026, we expect revenue of $130 million-$145 million, which at the midpoint equals a 47% increase year-over-year from the low end of our 2025 preliminary financial range. We expect public sector revenue to increase between 60%-70% year-over-year, with the rest of the growth coming from our commercial enterprise sector, mainly from VDR. Our Veritone Hire products and services are part of this growth, and we expect Veritone Hire to be flat to slightly down year-over-year, given the current macroeconomic hiring environment. Our managed services is expected to be up year-over-year by 10%-15%, chiefly because of the recent improvements we are seeing on the representation side of our business. We expect gross margins to vary between 60%-65%, driven by the forecasted revenue mix in the period, and non-GAAP net loss to be between $13.5 million-$22.5 million, which at the midpoint equals a 54% improvement year-over-year at the low end of our preliminary 2025 financial range. This adjustment reflects how revenue timing has shifted, the R&D increase we'd planned and already talked about, plus gross margin pressure stemming from the VDR mix in 2026. Our view is that profitability remains ahead of us, and the soonest that could arrive is Q4 2026, contingent to a large degree on VDR's anticipated compound growth across the public sector during fiscal 2026. I also want to call attention to a number of substantial balance sheet gains we've achieved. During Q4, we eliminated 100% of our senior secured term debt and bought back roughly 50% of the convertible notes that were outstanding at that time. As a result, our yearly debt carry costs fell 90%, from about $14 million down to only $800,000. Fiscal year 2025 ended with unencumbered cash and cash equivalents totaling $27.7 million, with no restricted covenants attached, alongside $45 million in 1.75% convertible debt and 92.6 million shares outstanding. Once our fiscal 2025 results are finalized, we intend to deliver a complete financial release covering Q4 and the full year 2025, and we anticipate furnishing it alongside our 2025 annual report on Form 10-K. That wraps up what I had prepared. Ryan?
Thanks, everybody, for making time for this today. In only a handful of quarters, Veritone has traveled quite a distance. Across the last 2 years, we've centered the company on AI, and that core platform underpins essentially our whole corporate and product offering strategy. The scale we're starting to see has us genuinely enthusiastic. Take VDR, for instance — it has grown into a genuine production engine, and we now work with each of the major hyperscalers, which is a huge achievement. Further evidence of our success comes from the strategic data set transaction and the Oracle partnership; those are widening our reach into high-value data while handing us the infrastructure needed to scale with greater efficiency down the road. Entering 2026, our attention is on disciplined scale: turning pipeline into revenue, growing our data supply, and keeping the momentum going on the operating leverage we've built. The platform, the partners, and the foundation are all in place; what remains is execution and scaling into what we see as a very sizable opportunity. Demand is still strong. Our pipeline keeps expanding, and our work with hyperscalers has never been stronger. I'm grateful for everyone's time today, and we can't wait to talk with you again next quarter. Thank you.
That brings us to the end of this call. We appreciate your joining us for today's presentation. You're welcome to disconnect at this point.
For developers and AI pipelines
Programmatic access to Veritone, Inc. earnings transcripts and 131,000+ others is available through the
EarningsAPI REST API and the hosted MCP server.
Quarterly plans from $162 - full transcripts, speaker segments, full-text search,
and the /api/v1/transcripts/recent polling endpoint for ETL pipelines.