管理層發言
Good day, and welcome to the AvePoint Inc. second quarter 2026 results conference call. I would now like to turn the conference over to Jamie Arestia, Head of Investor Relations.
Thank you, operator. Good afternoon, and welcome to AvePoint's second quarter 2026 earnings call. With me on the call this afternoon is Dr. Tianyi Jiang, Chief Executive Officer, and Jim Caci, Chief Financial Officer. After preliminary remarks, we will open the call for a question-and-answer session. Please note that this call will include forward-looking statements that involve risks and uncertainties that could cause actual results to differ materially from management's current expectations. We encourage you to review the safe harbor statements contained in our press release for a more complete description. All material in the webcast is the sole property and copyright of AvePoint, with all rights reserved. Please note this presentation describes certain non-GAAP measures, including non-GAAP gross profit, non-GAAP gross margin, non-GAAP operating income, and non-GAAP operating margin, which are not measures prepared in accordance with U.S. GAAP. The non-GAAP measures are included in this presentation as we believe they provide investors with a means of understanding how management evaluates the company's operating performance. These non-GAAP measures should not be considered in isolation from, as substitutes for, or superior to financial measures prepared in accordance with U.S. GAAP. A reconciliation of these measures to the most directly comparable GAAP financial measures is available in our second quarter 2026 earnings press release, as well as our updated investor presentation and financial tables, all of which are available on our investor relations website. With that, let me turn the call over to TJ.
Thank you, Jamie, and thank you to everyone joining us on the call today. Q2 was another strong quarter for AvePoint, and our results make clear that in today's rapidly evolving AI-dominated business landscape, our customer value proposition is resonating better than ever. This is true for organizations which are new to AvePoint, as well as for our existing customers, which are now consuming even more of the AvePoint Confidence Platform. This broad-based demand is reflected across our second quarter performance, where we once again outperformed all guided metrics, delivered double-digit growth in net new ARR, our 13th straight quarter, and achieved several record KPIs. As Jim and I will discuss, AvePoint has never been in a stronger position to capitalize on the multiple growth opportunities ahead of us. What is that value proposition for customers? This is where I want to spend my time today. Let's jump in. Agentic AI represents perhaps the largest technological opportunity of our generation, and companies around the world are moving swiftly to ensure that they fully harness its capabilities. In the course of these deployments, employees at organizations of all sizes, in all regions, and across every industry have moved beyond chatbots and Copilots and are now creating autonomous agents that can access sensitive data, make critical decisions, and take action across the enterprise. As these agents proliferate, new challenges are emerging, and we frequently see that many organizations cannot answer some very basic questions, such as how many AI agents are operating in the business and who owns them? What data are they accessing, and are they fully compliant with company policies? What risks are they introducing, and can you proactively recover from any granular damage they cause? What do these agents cost now and at scale? Perhaps most importantly, are these agents delivering measurable business value? Our customer conversations show that answering these simple questions is, in fact, far more difficult because enterprises today are facing a new form of sprawl. Across Microsoft 365, Google Workspace, Salesforce, Copilot Studio, and countless custom environments, AI agents are being created faster than organizations can manage them. As a result, we're seeing the emergence of shadow AI, agents operating outside traditional governance frameworks, often with access to sensitive information and little organizational oversight. Our formal research proves this as well. Our recently published State of AI report, which surveys 750 global IT leaders, found that 88% of organizations reported at least one security incident tied to agents in the past year, and that nearly 20% don't know whether their employees are using unsanctioned tools to build AI agents in their own environment. Without answers to those questions, without trust in this new, powerful technology, and without the ability to address shadow AI adoption and ROI at scale will eventually hit a wall. It's not surprising, therefore, that nearly 90% of organizations have delayed their AI deployments by an average of six months, citing this lack of trust. Just as cybersecurity became essential to the internet and governance became essential to the cloud, trust will become the essential unifying layer for AI, and this is where AvePoint is uniquely positioned. For 25 years, AvePoint has helped organizations govern, secure, and manage their most critical enterprise data. We have built deep expertise in understanding who has access to information, how data is being used, and how organizations can scale innovation without scaling risk. That foundation becomes even more valuable in the age of AI, because AI agents are only as trustworthy as the data, governance, and controls surrounding them. This is precisely why we launched AgentPulse, which became GA in Q1 as part of the tiered bundles of our control suite. At a high level, AgentPulse gives organizations unified visibility, governance, and cost control for every AI agent operating across their environment, all from one centralized command center. More specifically, AgentPulse can discover previously unknown agents, identify risky permissions, enforce data governance policies, retire unused agents, assign ownership, and gain visibility into the true economics of their AI investments. Critically, our approach is platform-neutral. We can do all these things across the most commonly used cloud ecosystems and business applications. Distinguishing AvePoint from the more narrowly focused point solutions, as well as the individual native capabilities of the hyperscalers. This is the value that only AvePoint can provide today, and we believe will drive incremental demand in the years to come. Gartner is currently defining a new agent management platform category, where it projects that by 2029, enterprise investments will exceed $15 billion, and the average Fortune 500 company will be managing 150,000 agents. This creates a powerful strategic position for us. As AI adoption grows, organizations will need more than basic intelligence. They will need the visibility, accountability, compliance, and operational control of the AvePoint Confidence Platform, the unifying trust layer for AI. Importantly, our ongoing innovation ensures that we will capitalize on this durable and growing market opportunity. This includes the launch of AgentPulse on a standalone basis in early July, as well as our announcement this week at Black Hat of new kinetic classification and rapid recovery intelligence capabilities. Two solutions which work hand-in-hand to continuously evaluate data sensitivity, provide security teams with a real-time view of their critical data, and help them quickly restore it at machine speed when an incident occurs. That's where we see the market going. Let me come back to the quarter and share some examples of the team's success with both new logos and existing customers. One of the largest American retail corporations has been a long-time AvePoint customer, and in Q2 opened up AI agent building to its 36,000 employees, immediately introducing risks around agent sprawl, cost exposure, and compliance. After evaluating native tools and other leading cybersecurity vendors, they chose AgentPulse through our Control Plus bundle, giving them a single pane of glass to immediately inventory, analyze, and govern more than 10,000 agents. By becoming the trusted layer beneath their entire agent environment, AvePoint is now critical to how they scale AI safely, deepening a relationship that now spans governance, security, and resilience across their organization, and elevating this customer into our $1 million ARR cohort. Similarly, one of the largest dental insurance providers in the United States, with more than 12,000 employees, faced a lack of visibility and potential sprawl as AI Copilots and agents began proliferating across their environment. After evaluating alternative solutions like Agent 365, they selected AgentPulse based on our more actionable governance capabilities and lower overall costs. This expansion successfully transitioned this existing customer from à la carte licensing to our Control Plus bundle, validating a land and expand motion that extends our trusted governance relationship directly into the management of their emerging AI ecosystem. Two Canadian corporations became new AvePoint customers in Q2. One, a consumer lender, knew that the native backup and archiving capabilities from Salesforce could not meet their long-term requirements. After successful proof of concept, the organization trusted AvePoint to protect their critical Salesforce data, and since that deployment, we have created new opportunities to expand into Microsoft 365 protection, along with broader governance capabilities from the control suite. Second, a financial services company came to us at a crossroads. Their new CISO paused a planned Copilot deployment until concerns around sensitive information could be addressed. We quickly acted on a proof concept, which provided the visibility and oversight their CISO required, removing a key barrier to AI deployment for this organization. With the confidence to move forward with Copilot across approximately 3,000 employees, the customer is now evaluating additional capabilities within our resilience suite. Lastly, one of the world's largest banks signed a seven-figure upsell deal in the quarter. A long-standing customer of all three of our suites. They needed to address strict legal requirements for record management and data protection ahead of a multi-petabyte modernization effort. As part of this expansion, the customer extended their governance and records management capabilities within our control suite to include both on-prem and cloud data. By doing so, the organization has further strengthened its governance framework and laid the foundation for secure and compliant AI adoption. AvePoint's mission has always been to give organizations the confidence to innovate faster while maintaining control of their data, security, and compliance obligations. That mission was relevant when we founded the company 25 years ago and is equally critical today as customers rely on us to safely deploy enterprise AI at scale. We believe the winners of the AI era will not simply be the companies building the most agents. The winners will be the companies enabling enterprises to trust those agents. As investors evaluate the long-term opportunity in artificial intelligence, we believe one theme will become increasingly clear. The future belongs to trusted AI, and AvePoint is building the unifying trust layer for AI that makes that future possible. Thank you again for joining us today. I will now turn it over to Jim.
Thanks, TJ, and good afternoon, everyone. Thanks for joining us today. Those of you who have followed the AvePoint story since our first Investor Day in 2023 know that the pursuit of our longer-term strategic priorities, as well as our quarterly results along the way, have been driven by a few key mantras. These include our unwavering commitment to profitable growth, our focus on controlling what we can control, and the importance of consistent execution and delivering on what we said we'll do. This mindset allowed us to achieve our longer-term goals of GAAP profitability and the Rule of 40 well ahead of schedule. It has also produced quarterly results consistently highlighted by outperformance on the top and bottom line, as well as steady improvements to key customer and operational metrics. We are pleased to report another set of these results today, as Q2 was highlighted by an acceleration of both total ARR growth and total revenue growth after adjusting for FX, record net new ARR dollars, and meaningful acceleration of net new ARR growth. Strong execution across verticals and customer segments, especially at the enterprise level, and our eighth straight quarter of GAAP operating profitability and the ongoing expansion of our GAAP operating margins, even as we continue making incremental strategic investments across the business. With that, let's dive in a bit deeper into the quarter. Total Q2 revenues were $124.5 million, representing 22% growth year-over-year and above the high end of our guidance. On a constant currency basis, total revenue growth accelerated to 21%. Q2 SaaS revenues were $98.5 million, growing 27% year-over-year and representing 79% of total revenues. On a constant currency basis, Q2 SaaS revenues grew 26% year-over-year. Term license and support revenue was flat year-over-year and represented 8% of Q2 revenues, compared to 10% a year ago. Lastly, services revenue were $15.7 million and represented 13% of Q2 revenues, compared to 14% a year ago. Turning to our revenue performance on a regional basis, in North America, total revenue growth accelerated to 23% year-over-year, driven by SaaS revenue growth of 27%. In EMEA, total revenue growth was 27% year-over-year, driven by SaaS revenue growth of 28%. In APAC, total revenues grew 16% year-over-year, driven by SaaS revenue growth of 27%. On a constant currency basis, EMEA SaaS revenues increased 26%, while total revenue growth accelerated to 24% year-over-year. For APAC, SaaS revenues increased 27% on a constant currency basis, while total revenues increased 16%. Switching to ARR, which we believe is the most important metric for investors, we saw strong performance from all three regions in Q2, as North America ARR grew 21%, EMEA ARR grew 34%, and APAC ARR grew 25%. Taken together, we ended the quarter with total ARR of $465.1 million. This represents 27% year-over-year growth and 24% after adjusting for FX, both of which are an acceleration from Q1. As a result, net new ARR in Q2 was a record $29.9 million, representing growth of 35% year-over-year and a meaningful acceleration from last quarter. Additionally, we are pleased that all three of our regions are now above $100 million in ARR, as our APAC business achieved this milestone in Q2. We ended the second quarter with 911 customers with ARR of over $100,000. This represents 26% growth, which is both an acceleration from Q1 and the highest growth for this metric in more than three years. More importantly, our larger customer cohorts of greater than $250,000, greater than $500,000, and greater than $1 million of ARR each grew at or above 30%. Finally, we are pleased to have added a record number of net new logos for the $250,000, $500,000 ARR cohorts in the quarter. Taken together, these results continue to show the durable and accelerating enterprise demand for our ability to solve their most critical data management challenges. This ability extends to every customer segment we serve, as we also added a record number of net new SMB logos in Q2. Much of this success is attributable to the continued rapid growth of our MSP business and strategy of driving more business through the channel. At the end of Q2, 59% of our total ARR came through the channel, compared to 56% a year ago. Two-thirds of our incremental ARR in Q2 came through the channel. Our MSP segment continues to be one of AvePoint's fastest-growing areas, and we intend to continue investing here to ensure we efficiently capture the enormous market opportunity that it serves. Turning now to our customer retention rates. Adjusted for the impact of FX, our Q2 gross retention rate was 89%, and our Q2 net retention was 110%, both of which were in line with Q1. Similar to prior quarters, our migration products again served as a two-point headwind to GRR, given their naturally lower retention rates. On a reported basis, Q2 GRR was 89% and NRR was 111%, both of which were also in line with Q1. Turning back to the income statement, Q2 gross profit was $91.7 million, representing a gross margin of 73.7%. This compares to 74.8% a year ago, and while the year-over-year decline is again the result of lower services gross margins, the 83% gross margins on our software products is in line with both the prior quarter and the prior year. Q2 operating expenses totaled $71.5 million, or 57% of revenues. This compares to 56% of revenues a year ago, reflecting our plan for increased investments across the business in 2026. As a result, Q2 non-GAAP operating income was $20.3 million, which represented an operating margin of 16.3% and was above the high end of our guidance. Importantly, we continue to focus on GAAP profitability through our ongoing management of stock-based compensation, which was 8% of Q2 revenues, compared to 11% a year ago. As a result, GAAP operating margins were 8.2% in Q2 and expanded nearly 130 basis points year-over-year and are now at 10% on a trailing 12-month basis. For the Rule of 40, which we define as the sum of ARR growth and non-GAAP operating margin, we finished Q2 at the Rule of 45 on a trailing 12-month basis. Using revenue growth and free cash flow margin to calculate the Rule of 40, we finished Q2 at the Rule of 47, again, on a trailing 12-month basis. Turning to the balance sheet and cash flow statement, we ended the quarter with $417.3 million in cash and cash equivalents. For the first six months of the year, operating cash flow was $40.2 million, or a 17% margin, while free cash flow was $37.7 million, or a 16% margin. This compares to operating cash flow of $20.8 million and free cash flow of $18.3 million in the same period a year ago. Lastly, we are pleased that on a trailing 12-month basis, our free cash flow has surpassed $100 million and is at 22% margin. Over the last few quarters, we have discussed the acceleration of our share repurchases, and last quarter noted that the $60 million that we utilized in Q1 outpaced the entirety of our buyback spend for all of 2025. This pace largely continued in Q2 as we spent approximately $50 million to repurchase another 4.9 million shares. When comparing our year-to-date buybacks with the first half of 2025, we have repurchased nearly 9x as many shares this year at approximately 70% of the cost per share. Through the close of trading on Friday, we have bought another 853,000 shares for approximately $10.7 million. Taken together, we have spent $121.5 million this year on share repurchases and have approximately $108.6 million remaining in our share repurchase program. These actions reflect our belief in the underlying strength of the business and our commitment to driving shareholder value. Importantly, the buybacks have more than offset the dilutive effects we see from our employee incentive programs. Turning now to our guidance, where I want to provide some color behind our current expectations. First, we are again raising our full-year guidance for ARR, reflecting our momentum and the demand we see. Second, similar to last quarter, our updated full-year guidance for revenue and non-GAAP operating income only includes the Q2 outperformance relative to guidance, as we account for the uncertain SaaS and term license revenue mix in the second half and the impact it may have on reported revenues. Third, given the enormous and rapidly growing market opportunity we currently see, we are increasing our expense plans for the second half of the year with two primary areas of focus. The first is technology, where existing investments have already driven rapid productivity improvements for our engineering teams and where further investment will accelerate the R&D transformation and drive similar efficiencies across the business. The second area of focus will be our go-to-market motion, where additional support for sales capacity, partner enablement, and enhanced brand awareness will allow us to better capture the market demand, support pipeline growth, and improve campaign conversions. Lastly, similar to last quarter, the final point is around FX, where the global nature of our business exposes us to fluctuations in currency exchange rates. Our updated guidance reflects the corresponding incremental FX headwinds we expect for the rest of the year, which more than offset the ARR raise and the Q2 outperformance. In other words, absent the impact of FX, our full-year expectations for top-line growth have accelerated relative to our guidance last quarter. As a result, for the third quarter, we expect total revenues of $128.2 million-$130.2 million, or growth of 18% at the midpoint. This includes an FX headwind of $2.4 million that is incremental to what was implied in our prior full-year revenue guidance. On a constant currency basis, we expect revenue growth of 19% at the midpoint. We expect non-GAAP operating income of $21 million-$22 million, and for the full year, we now expect total ARR of $522.1 million-$528.1 million, or growth of 26% at the midpoint. This includes a $1 million raise from our prior guidance, offset by an incremental FX headwind of $2 million. On an FX-adjusted basis, we expect total ARR growth of 20% at the midpoint, a modest acceleration from our prior guidance. We now expect total revenues of $508.5 million-$512.5 million, or growth of 22% at the midpoint. This includes the Q2 beat of $3.7 million, offset by an incremental FX headwind of $5.6 million, on a constant currency basis, we now expect revenue growth of 21% at the midpoint, an acceleration from our prior full-year expectations. Lastly, we now expect full-year non-GAAP operating income of $86.4 million-$88.4 million, which includes the Q2 beat of $1.2 million, the offsetting FX headwind of $1.9 million, and the elevated investments I discussed a moment ago. Finally, on a Rule of 40 basis, the midpoint of our updated full-year guidance is the rule of 43. We have again included a slide in our investor presentation that provides a walk from our prior full-year guidance in May to today's updated outlook. In summary, this was an excellent quarter from our team, with multiple data points confirming that our momentum and success with new and existing customers continues to strengthen. As we look to the market opportunity ahead of us, we see ample opportunities to continue capitalizing and driving shareholder value, both in the second half of 2026 and in the many years to come. Thanks for joining us today. With that, we would be happy to take your questions. Operator?
分析師問答
We will now begin the question and answer session. Our first question comes from Shrenik Kothari with Baird. Please go ahead.
Thanks for taking my question. TJ, it's really encouraging to see the early AgentPulse traction, particularly the wins you mentioned, customers choosing you for more actionable governance and lower total cost. You initially bundled AgentPulse within Control, and you have since launched it as standalone as well. Just curious how much of the pipeline uplift today is directly attributable to AgentPulse, and how are you thinking about standalone adoption versus using it to pull customers into the broader Control Plus deployments? Then a quick follow for Jim. Thanks.
Thanks for the question. We're very pleased with reception thus far on the AgentPulse product. It's definitely driving customer conversations and interest in the bundles. This is also why we announced the standalone in July, because the demand is quite high. The number of Control package customers roughly doubled in Q2 versus the prior quarter, and our pipeline statistics cited last quarter around Control bundles are roughly the same percentage so far this year. We did note that the average deal value with AgentPulse included is 2x-3x larger. Also, given the latest data, we see that on average, AgentPulse customers are managing well over 5,000 AI agents. What's remarkable is that the number of those agents is growing, doubling every quarter — every three months. This is definitely a high priority item for our customers, and this is why we made it a standalone SKU.
Great. Very helpful. Just a quick follow-up for Jim. I know you previously had disclosed Control represents roughly 40% of pipeline, and today's commentary suggests that the momentum has strengthened further. However, the NRR still remains and hovers around that 110%, has not yet begun moving towards that 115% target that you've set. If you can help just unpack a bit on the dynamics, like why are these larger expansions, as TJ just mentioned, including Control and AgentPulse, not yet translating into stronger expansion, and what should we expect next? Thanks.
Sure. Thanks, Shrenik. Maybe to keep in mind, it's still very early in terms of AgentPulse. As we mentioned in Q1, it was introduced midway through the quarter. We've seen really good pipeline creation and demand for those bundles that TJ referred to, but most of that is still in pipeline. We saw nice closings and bookings in Q2, but there's still a ton of pipeline. In addition to that, we just introduced the standalone SKU in July. We haven't seen any real impact on the Q2 numbers for that. We do expect to see an impact moving forward. We feel really good about the 110% NRR where we stand today. We still have that long-term target of 115%. We feel good about that. I do think over time we're going to see our NRR continue moving toward that 115%. Again, I feel good about the pipeline we've created and that the longer-term target will help us achieve the NRR goals that we've set.
Very helpful. Thanks a lot, TJ and Jim. Thanks.
Thank you.
Our next question comes from Joseph Gallo with Jefferies. Please go ahead.
Hey, guys. Thanks for the question. TJ, love all the prepared remarks on governance. That was certainly a key theme, Black Hat. Obviously, it's growing your pipeline. I wanted to ask, do you think you have the brand awareness in the market where you want it? Is that where the incremental OpEx spend is going? I'm just curious because I imagine it's a tremendously easy cross-sell, but can this also be a net new logo driver?
That's a great question. I'll go first, and then Jim can chime in. In our ecosystem, we have the brand name, especially within the Microsoft Cloud ecosystem, and we had a large team at the Black Hat Conference as well, and we highlighted new product releases this week around kinetic classification as well as rapid intelligence recovery capabilities. It's all part of our AI trust layer framing, and increasingly customers see our Confidence Platform as the trust layer for AI to confidently scale AI deployment without scaling risks. These messages are resonating with our customers and partners. In terms of branding outside of our ecosystem, we're going to do more. Jim mentioned specifically that we are increasing work on go-to-market and branding. We are investing for growth because we see the demand in the market, we see the need for our solutions, and we see actual evidence of growth for our product line. You will see more product announcements coming as the way to build out this AI trust layer across data infrastructure, AI, and AI agents. Stay tuned to see more of that from us.
Awesome. Maybe just as a follow-up for Jim, how was U.S. Fed demand? What's the pipeline look like? What are you expecting in the second half this year? Is that what's giving you the confidence in the ramp or the further acceleration in ARR growth constant currency? Thank you.
Thanks, Joe. I think you're spot on. We definitely see pipeline creation really accelerating, and that gives us confidence. Also, our ARR is following a trend we've seen over the past several years where Q2 accelerates over Q1; historically, our second half of the year accelerates from the first half of the year. Right now, 2026 is playing out exactly that way. We see nice pipeline acceleration from some of the things TJ mentioned about AgentPulse. That's a tailwind and is part of the reason that we see this demand and market opportunity, which is part of why we're doubling down and increasing our investment in the second half of the year. All of that taken together gives us confidence not only in the growth, but also the reason that we're increasing our investment.
Awesome. Thank you very much.
Thanks, Joe.
Our next question comes from Jason Ader with William Blair. Please go ahead.
Thanks. Good afternoon, guys. I was just wondering if you could help us understand how you're positioning versus Agent 365 from Microsoft, and just how to think about any potential headwinds from the E7 bundle where Agent 365 is included. Just help us think through that.
Thanks, Jason. We actually mentioned a customer example in the prepared remarks around how customers selected us. We can work hand-in-hand with Agent 365, but we also go one level deeper, similar to how we work with Purview. Purview is at the content level; we're at the workspace level. We provide delegate administration, cross-cloud capabilities, and tracking of Agent ROI and costs. We go beyond what the hyperscalers themselves offer, and more importantly, we provide multi-cloud capabilities. Agent 365 standalone is about $15 per user, and you need E7 at $99 per user to have it. Many customers have mixed license types; not everyone will move to E7. There's opportunity for customers and partners to leverage our multi-cloud management capabilities at an affordable rate.
Got you. Okay, thanks. Could you remind us of the pricing for AgentPulse?
There are two different flavors. In the bundle, it's embedded as part of the overall Control Suite bundle. The standalone that we released has two different flavors. The base flavor is providing visibility and observability so customers can see what's happening in their environment. The step-up, or the more advanced tier, allows them to actually take action and govern what's happening. Two different price points: a base entry-level price for visibility, and a second-tier price for governance. Both are competitively priced. We think they're aggressive to attract new customers and help existing customers take the next step. Early indications are very positive, and we're excited about what we'll see in the future.
All right. Thanks, guys. Good luck.
Thanks, Jason.
Our next question comes from Rudy Kessinger with D.A. Davidson. Please go ahead.
Hey, this is Ben Smith on for Rudy. Thanks for taking the question. Curious on the incremental investments that you're making. Do you see that as making investments to meet the ARR guidance for this year, or is that more intended to drive sustained growth in 2027 and beyond?
Thanks for the question. We broke those investments into two components: a technology-driven investment and go-to-market motion. Both have more future impact than immediate impact. The go-to-market investments are intended to capitalize on market demand and will have longer-term impact. The technology investments are already driving efficiency gains in engineering, and we expect continued transformation beyond engineering. Those efficiencies will happen over time. Think of these investments as longer-term; they may slightly impact current operating income, but they give us more confidence in hitting longer-term operating income targets because of the efficiencies and the ability to increase the top line.
Our next question comes from Erik Suppiger with B. Riley Securities. Please go ahead.
Two follow-up questions. One on the pricing for AgentPulse standalone: did you say that you price it on a per agent basis, or what is the foundation for that? Then on the operating margin, I'm assuming you're still targeting the 25%-30% in fiscal 2029. Can you describe what the trajectory for getting there will be? Will it be back-end loaded, or what should we be thinking about as you get past fiscal 2026?
Great question. For pricing, we've tried to make it easy for customers. Right now, it's priced on a per-seat or per-user basis, as opposed to a per-agent basis. Some customers have difficulty even understanding how many agents they have, so to provide pricing certainty, we chose the user model, and that seems to be resonating. When we think about operating margins and our long-term targets, we haven't changed those targets. The investments we're making today give us more confidence in achieving those longer-term targets because of the efficiencies we're seeing. The trajectory won't be completely linear; there will be ups and downs as we move toward the goal. We expect improvements over the years, but not at a constant rate. We're focused on GAAP profitability as well, and the investments today make the long-term goal easier to hit.
Just real quick on that user pricing: is that typically an administrator, like an M365 administrator, or who is the typical user that you're selling to there?
Think more like the Microsoft licensing model: it would be for the whole enterprise in terms of all users. That would be the structure, which is similar to pricing for other products as well.
Okay, thank you.
Thanks, Erik.
Our next question comes from Todd Weller with Stephens. Please go ahead.
Thanks for the question. Could you hit on the resilience business and talk about kind of the demand trends and drivers you're seeing there? Related to that, how do you see AI impacting and influencing that business, both from a capabilities perspective but also from a growth and opportunity perspective?
The resilience side is still growing robustly. From an AI perspective, AI risks are real. Our State of AI report shows 88% of companies had some AI security incident in the last 12 months. AI damage happens at machine speed, which highlights the need for robust resilience capabilities. We include Backup Express and Restore Express capabilities to restore environments quickly, in minutes. We give customers the ability to prioritize with recommendations on which workloads and functional areas to restore first — the concept of a minimal viable company in a massive outage. These scenarios are happening in real life; for example, geopolitical conflict previously drove spikes for resilience because of the need to restore production environments in another hyperscaler cloud quickly. AI is driving further need, and that's why we released new products at Black Hat leveraging AI to continuously classify and add rich context to datasets, while identifying restore prioritization.
TJ, just a follow-up to that. What are you seeing in terms of AI applications and workloads being backed up by AvePoint's resilience solutions?
We back up agents now as well. This includes IaaS and PaaS infrastructure running in compute clouds such as Azure, GCP, or AWS. We actively back up and restore those and can store them wherever the customer prefers — their own data centers, the same cloud, or a different hyperscaler for failover. That demand is present, and we do more of that today than ever before.
Great. Thank you.
Thank you.
Our next question comes from Joe Vandrick with Scotiabank. Please go ahead.
Thanks for the question. Maybe one for TJ. It seems like we're starting to see an acceleration in how quickly large enterprises are adopting AI. We've seen really strong results at Microsoft with Copilot adoption. You're starting to see it flow through numbers for other software companies like Datadog and Snowflake. I'd love to understand what trends you're seeing in your customer base and if this is ringing true for you. I know roughly half of AvePoint ARR is mid-market and SMB. What does that mean for adoption? Is it taking a little bit longer to show up for these smaller customers than the larger enterprises that you also serve? What does that mean for AvePoint?
The example we give is that on average AgentPulse customers discover they have at least 5,000 agents, and that number grows, doubling every three months in many cases. We have customers where we're managing hundreds of thousands of agents. The number of agents now far exceeds the number of employees. Many of these agents are lightweight workflow-type agents, not full virtual employees. Microsoft's Office Copilot adoption has been robust; it's now 30 million user seats, and GitHub Copilot is at 50 million. That's across enterprise, mid-market, and SMB. In Q2, we added a record number of SMB new logos as well as a record number of $250,000-$500,000 ARR cohort logos. We see success across segments. SMBs are not necessarily slower in deploying AI; this year deployments are more targeted and cost-conscious, with a focus on outcome-based AI deployment. That's why the AI trust layer is so important now.
Very helpful. Maybe one follow-up. You continue to improve and add functionality to the AvePoint Elements platform. Can you talk more about the traction there, and ultimately, how big can this MSP-focused business be for AvePoint?
Right now, MSP is almost analogous to SMB for us. SMB is roughly about 20% of our recurring revenue. We're adding capabilities to AvePoint Elements, including endpoint management, license management, SaaS license management, and agent cost management. It's our fastest-growing segment, and it continues to be. We think the SMB segment could be as large as 30%-40% of our business in the next few years. We still have a lot of greenfield opportunity: we have about 6,000 total partners and around 2,000 MSP partners, while the MSP ecosystem in North America alone is about 20,000 MSPs. There's a lot of room to grow.
Thank you.
Thanks, Joe.
Thanks, Joe.
Our next question comes from Nehal Chokshi with Northland Capital Markets. Please go ahead.
Thank you, and congrats on another good quarter. As always, great clarity on the guidance, including slide 25 that bridges the prior guidance and updated guidance. It shows that indeed it's a raise here, excluding the FX impacts. Just to be clear, though, this guidance change — is that reflective of or inclusive of Q2 outperformance on the ARR side?
We don't guide quarterly on ARR. We technically guide for the full year on ARR. We were pleased with the performance of ARR in the quarter, and that is reflected in our overall change to the guidance, where we feel confident based on Q2 performance and what we see for the rest of the year to raise guidance by $1 million. I wouldn't say it's specifically just about Q2. Thank you for recognizing the slide and the walk we put together. It can be complicated, and we appreciate that you noticed the detail.
Absolutely. TJ, there appears to be new ecosystems that have been born, specifically Anthropic and OpenAI. Do you agree with that assertion? If so, what are you doing to develop products specific to these new ecosystems and go-to-market mechanisms for these new ecosystems?
Those are commercially available large language model ecosystems, and they're increasingly important as organizations consider token costs and specific model capabilities. Customers use different providers, private models, and open-source models running on their infrastructure to be cost-conscious and outcome-driven. We remain model-agnostic: our focus is on foundational data — curating, adding context, kinetic classification, lifecycle management, and recertification. That improves outcomes regardless of which model the customer uses. We support backing up and restoring agents, including memories, models, and skills — the cached data repositories agents use to be productive. We're model-provider agnostic, and we support multiple ecosystems.
Could you give your perspective for enterprises using AgentPulse: what is the distribution of their agents across various LLM ecosystems — open source versus closed source versus hyperscaler-backed?
It's very dynamic and hard to pin down. Last year many focused on token usage as a proxy, but this year there's more cost consciousness. Enterprises are getting smarter about routing workloads and optimizing models. New open-source models and other commercial models emerge quickly, so snapshots become outdated fast. I don't think it's useful to provide a static distribution; the dynamics change by the quarter.
Thank you very much for that perspective.
Thanks, Nehal.
Our next question comes from Derrick Wood with TD Cowen. Please go ahead.
Great. Thanks. This is Cole on for Derrick. TJ, follow-up to an earlier question: with Mythos and all these new vulnerability concerns, are you seeing incremental demand and pull forward or faster buying cycles from customers?
Yes, that's what we're seeing with AI resilience. Damages are being done at machine speed, and recovery needs to happen at machine speed. That drives demand for resilience and the governance, the AI trust layer. Security is being disrupted by attackers that are now agents or fleets of agents. AI trust and AI recovery capabilities are increasingly important. You'll see more product from AvePoint in the coming months to build out that offering.
Great. Thank you.
Thanks, Cole.
Our next question comes from Kirk Materne with Evercore. Please go ahead.
Hi, this is Vinod on for Kirk. We realize this is a small part of the business right now, but could you talk about your Google Cloud related business and the growth trends there? Thank you.
We're pleased with Google growth rates. We've seen success in North America and Japan, and are now starting to see it in EMEA. Google is making good inroads into enterprise, and there are enterprise leaders who have moved there. It's a different tech stack, but the same problems to solve. We help customers move data between hyperscaler estates. The world is multi-cloud; many customers have both Google and Microsoft stacks. That area is growing well, and we're investing to ensure go-to-market success globally.
One more for me. The environment is changing quickly with Mythos, et cetera. Looking back to the end of last year, has anything surprised you or been different than initial expectations around the topic of shadow AI this year?
Things are moving very fast. CEOs are focused on two hair-on-fire problems: risk from shadow AI and agent sprawl, and cost — AI consumption costs are rising rapidly. Customers are concerned about budgets being blown by token consumption. These two pressures, risk and cost, are driving demand to deploy AI securely and cost-effectively. Nothing has surprised us in essence, but the pressure is higher this year versus last. Companies must control risk and cost while adopting AI.
Thank you.
This concludes our question and answer session. I would like to turn the conference back over to Tianyi Jiang for closing remarks.
Thank you for joining us today. I'll close by sharing where our confidence comes from, because I'm coming off a full round of quarterly business reviews with our teams across Americas, EMEA, and APAC. The message from every one of those conversations with our customers, partners, and leaders running each of our regions was remarkably consistent. The disruption created by agentic AI is real and accelerating, and organizations everywhere are turning to AvePoint to bring the visibility, governance, and above all, the trust this new era demands. We are executing well on every front, and our business is performing with greater breadth and consistency than ever. That is exactly why we're so bullish on our ability to capitalize on this generational opportunity as the unifying trust layer for AI, and why our conviction in the path to our $1 billion ARR target has never been stronger. Thank you again for joining us today. We look forward to speaking with you more this quarter.
The conference is now concluded. Thank you for attending today's presentation. You may now disconnect.