Good morning. My name is William. I'm with FiscalNote. Their CEO, Josh—Josh Resnik. Josh.
All right. Thanks a lot. Appreciate it. Appreciate everyone taking the time today to hear about what we're doing at FiscalNote. We're really excited and hope to be able to convey some of that to you and help you understand why. Let me make sure I can handle slides. Let's see. Yeah, not bad. I'll just start with the usual disclaimer, which I encourage you to read in full on our website. There may be forward-looking statements that we make today, and you should read all of our filings, including our 10-K, our 10-Q, which are also available on the website, and including risk factors. I want to first explain what do we do, and I'll talk about how people use our products and why it matters. We deliver subscription-based access to key, important policy and regulatory information. We provide core data.
We provide proprietary insights, and we also offer workflow tools and reporting tools. And it's through an AI-driven SaaS platform. This is really impactful information, that's important to, organizations of all types around the world. The reason why it's so important is you're seeing this increasingly complex web of regulation and policy globally down to locally. It used to be you could very carefully define, you know, who's a regulated entity, who's not, who needs to pay attention to this, and who doesn't. That's not really the case anymore. Even, you know, organizations that used to not be considered regulated per se are still, they're managing data that's regulated by privacy regulations globally. They're managing, all types of information flow. They're accessing global customer bases and the like. And so this is getting much, much harder for organizations to manage.
Because of that, the people who manage this, so the roles, which are things like government affairs, public affairs, legal, etc., are getting increased responsibility and visibility within their organizations. These are risks that large corporate boards are paying attention to now. That is certainly a tailwind for us as we think about how valuable this problem is that we solve for organizations. In addition, these organizations and these individuals are increasingly comfortable with leveraging technology and AI to solve these problems. That is helpful given how we do solve this for them. I'll walk through our product in just a moment. We see long-term the opportunity to do more and more to automate a lot of workflow that today is manual.
As generative AI capabilities improve, as agentic AI capabilities improve, we're very well positioned to take advantage of that and solve these huge problems for our end users in more and more impactful ways over time. These are all factors that are playing to our strengths and creating long-term opportunity for us as a company. Here you see some sample logos. It gives you a sense of the broad nature of organizations that work with us today. You see large corporates there on the left, and about half our revenue comes from corporates. In the middle is public sector government organizations, which include U.S. and global, and down to local as well. That's about a quarter of our revenue. Then nonprofits and NGOs, which is the remainder of the revenue.
Just to give you a sense for how they use us and why it matters to them. If you look at a company here like Nestlé, Nestlé obviously operates globally, but it starts with how they source ingredients. You know, it may be farming in South America. It may, you know, all sorts of—they source ingredients from many, many countries. They have to pay attention to regulations that impact them. It could be sustainable farming regulations, labor regulations, shipping regulations, etc. As they move these ingredients then around the world, they turn these ingredients then into food products and again, ship globally, sell globally in almost every country around the world. You can imagine the difficulty in managing to tariffs to global regulations all the way down to local in terms of where they're sourcing this, where they're selling these products, etc.
It's a very hard problem for a company like that to solve. In terms of public sector within the U.S., we have many, many use cases. We serve all branches of federal government. We serve interstate and local as well. The use cases involve how these organizations are tracking their own funding. So eight federal agencies, how they think about funding that may be coming from Congress. It can include federal wanting to understand regulations that are happening at a local level and vice versa. Globally, different global government organizations need to track how others are regulating various issues. Again, it's an equally complex problem for public sector as it is in the private sector. For nonprofits and NGOs, it's these types of use cases. Plus, quite often they wanna activate their constituencies to influence lawmakers as well. Grassroots advocacy. And we have grassroots advocacy.
We have a Grassroots Advocacy product that they can use as well. That just gives you a sense for how these organizations are leveraging us to manage this type of issue. Here's a quick look at our customers. As I mentioned, it's split about 50% private sector, a quarter public sector, and a quarter nonprofits and NGOs. The typical end user is someone who has a title that relates to legal, government affairs, public affairs, and the like. As I mentioned, these are roles that are becoming increasingly important as these issues are recognized as impacting business outcomes for these organizations. We have about half the Fortune 100 and 92% of our revenue is subscription revenue. What do we actually do in order to provide this service? It starts with collecting this data.
The data that we collect, a lot of it is public data. You know, anything that's related to legislation and regulation is gonna be, you know, by definition, public data. The datasets are really difficult to get to. They're fragmented. They're unstructured. They're very difficult to find. They're very difficult to get. The data's very difficult to clean, etc. It's a very difficult problem to solve in terms of getting these datasets. We source from tens of thousands of sources in order to get the information that we pull together. We combine this with proprietary analysis that we create. We have human-level intelligence on top of the data that we provide. We have teams in Washington, D.C., in Brussels, and elsewhere.
Just to give you a sense for how special and unique it is what we do, our teams in D.C. are in the halls of Congress every day, tracking what's happening down to the finest detail. As an example, they'll be in the room as a congressional subcommittee is marking up a draft piece of legislation. They will be tracking which clauses get struck in the course of that, you know, because that's the sort of thing that doesn't get reported in major news outlets, but is incredibly valuable and useful information for our customers who need to understand how bills progress or how regulations are being developed. Our teams understand and are able to convey not just what's happening, but what's likely to happen, what's the politics behind it, what's the likelihood of it.
We have incredible knowledge and insights that we're able to provide. This is proprietary information that users can only find on our platform. We combine these very difficult-to-get public datasets with our proprietary analysis, and we leverage that and we provide insights. We provide workflow tools. We provide reporting tools on top of it. As I said, the public datasets are very difficult to get and to clean. You, as I said, it's tens of thousands of sources. The formats are all different. It is multimodal. It can be text, audio, video. It can be structured, unstructured. The sources are constantly changing as well. This problem of getting to the public datasets and actually making the data useful is really, really hard to do. We have spent a decade doing it. We differentiate through this combination.
We're not just an aggregation of public data. We're not just providing news and information. We're providing this combination of very difficult-to-get, core data with human intelligence on top of it. It's a mix of, kind of the scaled automated provision of information with the ability to also provide custom curation for customers who need it. We have a lot of large organizations in particular who really benefit from getting custom curation as they have very specific needs that they need to fill on a global scale. We can do that for them in a very economical way. It's a very unique and powerful value proposition that we bring. In terms of understanding where we are as a company, I'll walk through that now.
In the context of that, I'll walk you through the product in more detail so that you can get a deeper understanding of that as well and how it ties back to where the growth opportunities lie. The first thing, the three pillars that we're really focused on right now are adjusted EBITDA profitability, managing our debt and accelerating the path to positive free cash flow, and driving to resilient and durable long-term growth. To hit on each of those. Profitability, we have made a significant change in how we operate as a company over the past couple of years. We've gone from a significant adjusted EBITDA loss in 2022, where we lost $24.5 million, to positive $9.8 million last year. A huge swing in terms of how we operate. We're now consistently profitable, with seven consecutive quarters of adjusted EBITDA profitability.
That profitability continues to expand. Our margins this year will be double last year on a pro forma basis. We have proven that we can operate profitably to do it consistently and to continue to expand our margins. We have done this through very disciplined operations. We have done this through streamlining of management layers. We have done this through making sure that we are focusing our teams on the areas that are most primed for growth. The second pillar here is on reducing the debt and getting to positive free cash flow. Again, we have been very focused on debt management. We have reduced our senior term loan by 60% in the past year and a half. Our free cash flow is improving rapidly, and this has come through divestitures of non-core assets.
Again, that has the benefit of helping us to focus as a business in this area of policy and regulation, which we see as poised for long-term growth. It all continues to feed this virtuous cycle where we continue to improve profitability, we continue to reduce our debt, and that in turn enables us to focus where we need to, to drive to growth. That is the last pillar that you see here, which is, getting to this profitable, durable growth. We have a strong foundation for this. We have thousands of customers who use us today. They need the data and information that we can provide. We're seeing increased demand, as you can expect, in this environment where policy is changing so rapidly. The challenge that we've had from a growth perspective has really been the product, or in our case, products.
We have this wealth of data, that no one else has. We collect from more than 100 countries globally. We go deeper locally, than anyone else with more than 16,000 districts across the United States. This incredible wealth of data is incredibly valuable and important to our users. We combine it with our proprietary intelligence. The challenge has been that for users to access it in the past, they have had to access it through legacy siloed platforms. Federal data in one place, state data in a different product, etc. That creates a lot of friction in the selling process. That creates retention issues as users do not have the right engagement from day one.
That creates challenges for expansion revenue where cross-sell, upsell becomes a phone call from your account manager to sell you a new product as opposed to just adding datasets into your current experience. The legacy siloed products have been the single biggest challenge that we've had from a growth perspective. We are addressing that through launching a new platform that will consolidate all of our data and information in one place. That platform is PolicyNote, which we announced in January of this year. PolicyNote is a very much improved user experience, modernized experience, very AI-forward. Natural and intuitive ways to find your information.
Instead of configuring, say, a complex Boolean search, you know, I'm looking for this word within 30 words of that word, but not that word, you can just ask our AI assistant, "Hey, show me all of the legislation currently pending in the U.S. related to ethanol or whatever your issue or area of concern might be." Much easier to get in, much easier to start making use of the platform, and much quicker to get value out of it. You can see here a simple example of why this matters, with our AI alerting on the left. On the right, you have some customer testimonials. Just to talk through the alerting for a moment and why this intuitive experience really matters.
If you can imagine who our users are, if you picture yourself, say, as running government affairs or public affairs for a large multinational, and you need to report to your board on where your areas of highest risk are from a policy and regulation standpoint, the thing you live in fear of is missing something that you should have known about and not being aware that it was lurking around the corner and potentially creating significant impact for your business. By leveraging AI rather than legacy ways of finding this information, we're able to provide much more powerful tools, tools to our end users.
Here on the left, you see an example of our AI alerting where we're able to alert users not just to something that they said they need the keyword for, but based on the issues that we know are of concern to them, we're able to find more information that may be more relevant to them that may not be tied to a specific keyword. Here you see an example of something that would impact someone who's interested in ethanol production, but doesn't actually mention ethanol itself. That is really important for our end users at the end of the day. Again, you can see along the right some of the customer testimonials that we've heard in relation to this. We know that PolicyNote is working. We know that we're seeing now levels of engagement that we expect will drive that growth going forward.
Here's, in addition to qualitative testimonials that we've heard and that you saw on the prior slide, here you see some data that shows you quantitatively how we think about it. On the left, you see, we migrated prior to announcing PolicyNote, we were testing and we took a cohort of highly at-risk customers. Users who, based on their levels of activity or inactivity, we were very confident were likely to churn. We migrated them on a PolicyNote to see what difference we could have with the platform on their activity. You see that three quarters of those at-risk customers actually became engaged following that migration of PolicyNote. A third of them are what we consider to be power users based on their levels of activity.
We're seeing that we can take highly at-risk customers, customers likely to churn, and we can drive the levels of activity that will drive retention and growth. We obviously look at metrics much deeper than this. We look at engagement metrics across the board. We're seeing the levels of activity that we want to see. In terms of things like how frequently are users searching for information on the platform? How effective is the platform at providing them the information that they're looking for? How often are they using the AI tools to find insights? We look at those metrics very closely, and those are all highly positive, so far in excess of activity levels that we see in our legacy products and either exceeding or meeting what we consider to be the right industry benchmarks as well.
We're feeling very good about the levels of activity. We want to see our users do things like set up projects, set up alerts, all these things. We're seeing in PolicyNote, they're doing this at levels that are much, much quicker and more productive than what we see in the legacy platforms as well. All the indicators are in the right direction for PolicyNote. We're also seeing customers vote with their wallets. When we look at the share of ARR on new logo from corporate customers, we're seeing significant increases. You see this number on the right where it's more than double what we saw year over year. We're seeing customers who are coming in and committing to us for multiple years.
That's a strong indicator of confidence and obviously just mathematically has a direct impact on gross retention next year as well. The impact will be significant. We see that PolicyNote will drive benefits throughout the customer lifecycle. From a new logo perspective, we believe that we'll be able to drive ACVs higher from the start through better bundling. As I mentioned before in the context of cross-sell, upsell, having data siloed on legacy platforms introduces significant risk or significant friction when it comes to trying to create these bundles. By having all the data and information in one place in PolicyNote, we'll be able to do more bundling upfront and drive higher ACVs. We'll also be able to have more of a product-led sales model, which should improve all of our deal velocity metrics as well.
From a gross retention perspective, we'll see higher, we expect to see higher engagement. We're seeing that already on the platform, a better setup experience, which we have on the platform. As I said, that's driving higher levels of engagement. That should result in better retention, gross retention over time as well. Then expansion revenue through net retention, our ability to do cross-sell, upsell. We're reducing friction there as well. We'll be able to introduce not just new datasets to users to expand ACVs, but new features over time. Tied to that is we've totally changed how we operate from a product and engineering perspective. We've changed product and engineering leadership over the course of the past nine months. That's having a significant impact not just on how PolicyNote works today, but also the pace of product innovation generally.
When you have the right product innovation, that really solves two problems. One is it gives your customers and prospects a high degree of confidence in terms of their willingness to invest in your product. Again, when you think about customers who are signing with us for multi-year commitments where they would not in the past, that is a sign of that confidence where it is less even about what features you have today, but more, do I expect your platform to continue to meet my needs over time? That pace of product innovation. The other thing that that pace does is it enables you to introduce new features over time that not only can drive new engagement, but again, you can charge for some of these new features over time and continue to expand ACVs.
As an example of that pace of innovation, we've launched, since announcing PolicyNote in January, 15 major new product features and many more enhancements since then. It is this continual launch of new features and enhancements that continues to drive that engagement and that confidence and, over time, will help us expand ACVs. We are very happy with the pace of innovation and what we are doing there. To give you an example of how we have done that and how that has tangible commercial benefits, I am going to talk about the tariff tracker that we launched in April. There was the massive announcement of sweeping tariffs on April 2 when President Trump announced global tariffs on what he called Liberation Day.
That then led organizations to need to, respond and understand how this might impact their business, be able to track to what these tariffs meant country by country and the like. Within two weeks, April 2nd to April 16th, we launched a global tariff tracker inside PolicyNote. Just to pause on that for a moment, the pace of our being able to do that and to innovate in that way, that's pretty incredible to go two weeks from not having it at all to it being live available in the product. That in itself is valuable. There's also, you know, I talk a lot about how we execute and, and the improvement in our operations.
There is also how this ties back to where we are from a go-to-market standpoint and the close tie of being able to take what we do from a product perspective and tie it to commercial success. On April 16 alone, we added $1 million in new sales pipeline driven by the launch of this tariff tracker. To be able to execute to that degree of precision, to be able to move that quickly and improve pipeline in that way, that speaks a lot about what we are able to do going forward and how we are able not just to innovate from a product perspective, but to turn that into commercial success and growth. All right, here you see some of the numbers for 2025.
We've recently reaffirmed our guidance for the year, which is $94 million-$100 million from a revenue perspective, adjusted EBITDA of $10 million-$12 million. As I mentioned, the adjusted EBITDA margins are double, 2024 on a pro forma basis. I do want to speak for a moment about AI generally, because we often get the question of, you know, how much is that a threat versus an opportunity for us as a business? You know, will ChatGPT and the like eat us alive? We view AI as an accelerant for our business, and I'm very excited about the opportunities that it presents for what we can do in the near term and the long term. First thing that you need to understand about that is the importance of trust and accuracy to our customers.
Again, put yourself in the shoes of, say you're again, the head of global government affairs for a large multinational corporation. You need to report to your board. You can't report to them based on, you know, a ChatGPT result any more than you could tell them that you Googled something and got to a result, right? You need a system that you can rely on where you know you have accuracy of data, where you know the insights, you can trust and value them. That's what they get from us. We have a tremendous brand equity in this space. They know that they can trust the information that we provide. We're very, we're very protective of that trust as well.
As we go through testing our AI features, as we go through running our prompt engineering, we're very conscious of the need to drive clear and accurate information for our customers above all, and we won't take a risk with regards to that trust. How are we then able to guarantee that accuracy? How are we able to kind of fulfill that promise and that trust? Number one, you have to remember there's a differentiator here where we have essentially comprehensive data sets globally down to locally. Even if you assume that large other large language models like, you know, ChatGPT, you know, have the same information, we're running our queries only against those data sets that we know are relevant, right? This is the information that we know our customers care about. We're not running against the entirety of the world of information and policy.
We're not running against someone's blog who may be talking about regulatory issues or policy issues or like. We're running against the data sets that matter, plus our own proprietary analysis, which no one else has, but which we also know is trusted and accurate. We're running against a unique data set that's known to be trusted and accurate. That's a significant part of why we're able to fulfill this promise of this trust and ensure that our customers are getting accurate information. Why do we consider it to be an accelerant versus just having kind of a moat? We're leveraging LLMs in the background. We use OpenAI, for example, in the background as we run queries.
As those models get better and as generative AI and agentic AI capabilities get better, as they improve, our results and our capabilities improve as well. We see a tremendous opportunity in this space to leverage generative AI and agentic AI to solve more and more complex workflow problems for our users, which will enable us to take ACVs much, much higher over time. I'll give you a simple example of how we do that today. In the PolicyNote platform, if you're tracking a piece of legislation, then you probably also need to draft a position statement for your organization about that legislation. That today, if you do it manually, is probably days of work for you or your team. With PolicyNote, you can leverage our AI assistant to do that for you with the click of a button, right?
Leveraging our generative AI to do it. That's a relatively simple use case, but the world of what we can do here with generative AI as it improves, with agentic AI, we're able to do more and more automation of workflow that today is manual. If you think about our users, they spend probably a hundred times what they spend with us doing manual work with law firms, with lobbying firms, with public affairs firms and the like. If we can start to bring some more of that work into the platform and automate more of that workflow, that will enable us to grow that share of wallet, take our ACVs up significantly, and drive tremendous ROI for our end users. These were our Q1 numbers, where we exceeded expectations on both a revenue and adjusted EBITDA perspective.
I want to really just get to the key takeaways and then I'll see if anyone here has any questions. The first thing to remember is this focus in terms of product being this lever for us to drive this next phase of growth. We're shifting to this model of product-led sales and product-led growth, and we're really excited with what we're seeing in PolicyNote, the user engagement in PolicyNote, and our ability to drive that growth going forward. We're going to continue to exercise the operational discipline that has enabled us to be consistently profitable. We see continued opportunity to drive margin expansion. We see continued opportunity to get more and more efficient, especially as we drive efficiency in the areas of product and data operations. We're going to continue on that path to positive free cash flow.
So again, we, you know, we haven't guided to anything on positive free cash flow, but you can see the progress that we've made and understand how focused we are, on getting there. We're focused on delevering the balance sheet. We have paid down senior term loan significantly, and we're focused on managing our debt load over time, as we work towards that future growth. As I said, we've reaffirmed our 2025 guidance, and we expect to see ARR growth resume in the second half of this year as PolicyNote starts to take root and then continue to see that growth expand over time. With that, I'll stop there and see if we have any questions. Yes. Okay, so the question was about major competitors. There's a handful of others who are in this space. Two are Bloomberg and Politico.
They each have groups that do this. In both cases, a lot of the content tends to be an outgrowth of their news operations. They tend to be U.S. focused. In the case of Bloomberg, mostly federal. In the case of Politico, also mostly federal, some state. Politico does a little bit in Europe, but they're not comprehensive global to local the way that we are. They're not as data focused as we are. There's another company, Quorum, that has a platform that's more similar to ours. They do have global to local. They're not quite as broad globally and not as deep locally as we are.
They do have a consolidated platform, but what they do not have is the human intelligence that we provide on top of it, which, as I said, is very differentiated and very valuable to our end users at the end of the day. They do not have the same abilities to do custom curation and the level of customer support that we do for our customers as well. Any other questions? Yes. Yeah. The question was about our numbers around net retention. Most recently reported was in the low 90s, like 90%, 92%, 93%, in that range. Historically, we have been higher than that, and there is no reason that we, so we do not guide on net retention, to be clear, but over time, there is no reason that we should not be back to higher levels over time.
It shouldn't be, there's no reason that we shouldn't be at levels that you would typically expect to see from successful information services businesses. The biggest challenge that we've had from a retention perspective has been around the product because it drives, the challenge that we've had with our legacy siloed platforms has been driving the levels of user engagement that would naturally drive successful gross retention metrics and then the friction that's involved in the upsell cross-sell because of the legacy siloed products. With PolicyNote and with the success we're seeing on engagement, we expect to see those NRR numbers return to more expected levels over time. Yeah, we don't break out between the net and gross retention today. Those are the two major factors behind where we are.
Yes, we had a question. Yeah, sure.
The question was about the nature of proprietary data versus public data at the end of the day. In terms of the quantum, most of the data that we provide to our end users is our public data sets. You picture global legislation and regulation, picture U.S. federal, state level, down to local levels. We even have down to things like school board minutes that get published as well. Technically, that's public data. I mean, it's, you know, government produced data and it's public. There's a few things to think about. One is there's a challenge in finding all this data, getting all this data. It's easier said than done. Yes, it's public, but it's hard to get, it's hard to clean, it's hard to make it useful and be able to provide the insights to end users at the end of the day.
Then a second piece to think about is there are cases where we're able to provide this data better and quicker than others, in some cases because of the access that we have. Because we have these people, for example, who are in D.C. and in the room as things are being marked up, we sometimes get information before it's published through official channels. Sometimes there will be hard copies of documents that technically are public, but sit in one place and we're able to get a copy and get it into our platform before it's available elsewhere. Those types of things, getting that information quickly really matters to our end users at the end of the day. There's significant value in that speed to market, and then the insights that we provide are truly proprietary to us, right?
Those are that we produce, the analysis, the insights, et cetera, that our teams produce. Nobody else has that. Yeah, it's the question really about licensing that data to LLMs. Is that where you're going with it? I think the question fundamentally is, is there an opportunity or would we be licensing that data to LLMs? I think that data is valuable and useful. Again, it's very difficult to get. It's multimodal, which has traditionally been hard for them to get to and manage, et cetera. That being said, again, the proprietary analysis that we have is very unique. We have a lot of historical data as well. Some of our data goes back 80 years. The data and analysis that we provide, I believe there's a lot of significant value in that.
We believe, though, that we can leverage the strengths that are coming from out of the LLMs in terms of the capabilities that we bring and leverage our kind of unique aggregation of the public data, combine that with our proprietary information and kind of get the best of both worlds where we're not going to be disintermediated and we're able to provide more and more powerful solutions to our end users. Good. Do we have time? Can I take one more or? Okay, we can connect separately. But thank you all for your time. Thank you.