Thomson Reuters : Transcript (55355c)

REFINITIV STREETEVENTS

EDITED TRANSCRIPT

TRI.TO - Thomson Reuters Corp at JPMorgan Ultimate Services Investor

Conference

EVENT DATE/TIME: NOVEMBER 18, 2025 / 4:00PM GMT

CORPORATE PARTICIPANTS Steve Hasker Thomson Reuters Corp - President, Chief Executive Officer, Director Michael Eastwood Thomson Reuters Corp - Chief Financial Officer Gary Bisbee Thomson Reuters Corp - Head of Investor Relations CONFERENCE CALL PARTICIPANTS Andrew Steinerman JPMorgan Chase & Co - Analyst ‌PRESENTATION Andrew Steinerman - JPMorgan Chase & Co - Analyst

All right. I think that's a go time here. You're on the info services track of the Ultimate Services Investor Conference, our primer, the info services data book, which we've been updating quarterly since 2013 is up here. You can take it on your way out.

This is the Thomson Reuters discussion. With me today is Steve Hasker, CEO; CFO, Mike Eastwood; and IR, Gary Bisbee. I'm going to be asking a series of questions. We're together for about 30 minutes. I'm going to ask something like 20, 25 minutes of questions and then ask for your questions.

QUESTIONS AND ANSWERS Andrew Steinerman - JPMorgan Chase & Co - Analyst

‌I would say the biggest remarkable thing in -- since you took over as CEO has been the product road map. The product road map is just incredibly picked up. It's not just generative AI, but currently, lots of the enablement is generative AI. My question is, how do you know if you're doing enough product innovation? And talk specifically about your agentic products in Legal and Tax, and how client adoption has been?

Steve Hasker - Thomson Reuters Corp - President, Chief Executive Officer, Director

Yes. Thanks, Andrew. Thanks for having us. Just quickly. We're joined today by our Chairman, David Thomson. We have a Board meeting here in New York this afternoon and tomorrow. So welcome to David and also the CFO of our Corporate Business, Erin Brown, here.

So how do we know if the product innovate -- the rate of innovation is enough? Well, I would give you a simple answer that my colleagues will, sort of, cause some eye rolling, which is, it's never enough. And the reason I say that is that we serve -- we're lucky enough to serve, I think, the most incredible professions, the Legal profession, the tax accounting audit profession.

And, to a lesser extent, but an important extent the sort of risk, fraud and compliance and law enforcement professions. And the opportunities that AI presents to those professions, and the opportunities that AI presents to Thomson Reuters to play a larger role in the success of our customers is the likes of which I haven't seen in my career. And so we cannot innovate fast enough. We cannot improve CoCounsel and Westlaw and Practical Law, ONESOURCE, CLEAR. You name one of our product franchises and we have very aggressive road maps in terms of injecting agentic AI capabilities into those products, firstly, integrating them amongst the suite for TR, and then integrating them within our customers' own capabilities and data.

And so yes, Andrew, the rate -- we're proud of the rate at which the product innovation has improved over the last number of years. What I'm particularly happy with is a couple of launches that we put in place that we made at ILTACON, which is the big Legal tech conference in August. So CoCounsel Legal, which is the integration of the products and a new version of CoCounsel, and then Westlaw Advantage, which is the latest agentic and deep research version of Westlaw. In tax accounting, we have some very important launches coming up in the next couple of months.

And so we're excited about those. The early testing looks really positive. But the short answer to your question is it's never enough, and I hope that Mike and Gary will agree.

‌Michael Eastwood - Thomson Reuters Corp - Chief Financial Officer

Yes. Steve, just an additional point of Andrew's question was on client adoption. We're certainly pleased thus far. I think Andrew, we've discussed the investments that we've made the last two years in customer success, certainly driving the additional adoption and then the sustained usage. We have the customer success in each of the Big 3 segments, which is helping us monitor the progress --

Andrew Steinerman - JPMorgan Chase & Co - Analyst

[Favorite part] is customer success.

Michael Eastwood - Thomson Reuters Corp - Chief Financial Officer

‌Yes. Customer success is a group of individuals who help -- once we make a sell, they're working directly with our customers on any questions that they have. They help with the training, and then they monitor the adoption throughout the process.

Gary Bisbee - Thomson Reuters Corp - Head of Investor Relations

And maybe, Andrew, I'll just add one thought, which is if you take a historical perspective, a lot of what Steve and Mike did in their first couple of years, running the company really set up, this accelerating innovation. And so there was a clear focus on prioritization, investing more, understanding our franchises, investing more behind the best franchises and the ones with the longest growth runway. I think also through the Change Program, modernizing and streamlining our technology infrastructure really allowed us to move fast when generative AI came on the scene. And certainly, the legacy of talent we had in our labs and data science area also has been really helpful. And we've invested heavily, as you heard us talk about.

So I think it's important to note this isn't sort of -- this isn't sort of luck or a one-off opportunity in our markets. But it's been a real focus on growing our investment, prioritizing the best opportunities and moving quickly relative to the history of TR.

Michael Eastwood - Thomson Reuters Corp - Chief Financial Officer

Andrew, just a couple of points that were mentioned this morning and some of our breakout sessions. The common question has been the level of investment in GenAI. I stated at the beginning of this year, it would be north of $200 million, that holds. For full year 2025, it will be in excess of

$200 million. That's split roughly 50% operating expense, 50% CapEx. That level of investment will continue into 2026 and is factored into our guidance for both years.

Andrew Steinerman - JPMorgan Chase & Co - Analyst

Okay. Great. I was not at ILTA, but I do annually go to a Legalweek, which is in New York in February, or March. And maybe you're going to say a lot of time has passed since last Legalweek, but I didn't find the kind of user interface on Westlaw to be fully integrated like this, different, and this is really for corporates. There's just different stations there, different logins. How focused is Westlaw on kind of a smooth, kind of modern user interface experience? Because I know you're going to say like maybe going forward, you're not a module approach, but history is a module approach.

Steve Hasker - Thomson Reuters Corp - President, Chief Executive Officer, Director

Yes. So that was really the big -- one of the big breakthroughs at ILTACON was the sort of seamless integration under the CoCounsel legal brand of Westlaw Advantage, of Practical Law AI and of the CoCounsel legal assistant. So we've addressed that one, Andrew. And what I would say is we're pushing in two places. The first is to make sure that our user experience is as seamless as it possibly can be and appeals not only to the senior most practitioners, but also the entry-level lawyers.

And then secondly, that our products are significantly more accurate than the next best alternative in terms of the output that they produce and therefore, the time that they save.

Andrew Steinerman - JPMorgan Chase & Co - Analyst

And do you see that already in your Net Promoter Scores?

Steve Hasker - Thomson Reuters Corp - President, Chief Executive Officer, Director

Yes. So we've roughly doubled our NPS since we launched the Change Program back in 2021. And so it took a while to get there. I think some of us were a bit frustrated at the rate of improvement. But we started well below the sort of B2B averages, and we're now comfortably above, and we're just going to keep hammering that proposition because we think the extent to which we do that, it better differentiates the broad experience that customers have with Thomson Reuters. It allows us to get more price in order to see our cross-sell and our NRR improve in really healthy ways.

Andrew Steinerman - JPMorgan Chase & Co - Analyst

You guys, sort of, bravely have already talked about 2026 targets, specifically for the Big 3. You're targeting 9.5% revenue growth. Are you able to break that down at all for us, the '26 number, price, upsell, new customers?

Steve Hasker - Thomson Reuters Corp - President, Chief Executive Officer, Director

Gary, you'd like to start?

Gary Bisbee - Thomson Reuters Corp - Head of Investor Relations

Yes, I can take a cut at that. So Andrew, the last couple of years, and I think we would expect this to hold in 2026. We've been, for price, in the 3% to 3.5% range. We -- with some of our newer innovation, there are a few areas like specifically in Westlaw, with Westlaw Advantage, where we're pushing for a bit more price than we've had historically, but that will take time to flow through. So 3% to 3.5% price. I think the rest --

Andrew Steinerman - JPMorgan Chase & Co - Analyst

And that's enterprise, right? Or that's Big 3, right?

Gary Bisbee - Thomson Reuters Corp - Head of Investor Relations

Yes. I think it's total company, but I think it's also a similar number, Big 3. I think the rest of it is split into volume, which would be adoption products where we're driving more adoption and then mix. And mix, in my mind is sort of an upgrade, selling them a higher priced, higher capability version of a product. And so we've got growth in all three buckets.

Pricing across the board. There's a lot of products that are driving that real volume growth. In Legal, that's CoCounsel, that's Drafting, HighQ, Practical Law are still volume growers. In Tax & Accounting, there's a number of products SafeSend, the acquisition earlier this year, the GenAI launches of CoCounsel for tax and Ready to Advise, et cetera, incorporates Indirect Tax. Pagero, e-invoicing and others.

And so we've got a lot of volume. There's also, obviously, the big upgrade cycle in Westlaw remains a real driver. And by launching Advantage this summer, I think it's safe to say we've extended the runway of time over which we can continue to drive upgrades at Westlaw, and there's a number of other products across the estate where it's more of a mix. But I think it's pretty evenly balanced if you look at the growth in aggregate between those 3. Price, volume and mix.

Michael Eastwood - Thomson Reuters Corp - Chief Financial Officer

Gary, I'll just supplement, Gary mentioned Pagero and Indirect Tax. In aggregate, Pagero and Indirect Tax is growing in excess of 20% organically this year. Another point of emphasis is international. International, as a reminder, drives roughly 1/3 of our total organic growth. That's growing at 16% year-to-date. We would expect that to continue for the remainder of this year, and continue to accelerate as we go into 2026.

Andrew Steinerman - JPMorgan Chase & Co - Analyst

Most people in the room know that we did this chart. It's on page 8 of the information services data book about the AI continuum. And I really think when we were talking about Legal, I think of really two pieces to Legal at Westlaw. The US research library, which Steve, you did an amazing job of discussing the fortified moat there on the third quarter conference call.

I wanted to -- either one, you could truly maybe give a little recap there because I thought that was amazing, and fully agree with Steve. I wanted to talk about some of the competition that's in kind of Legal workflows, and really like spots that Westlaw is just not going to be in. Spots where, obviously within workflow, they're going to win. So maybe do you want to maybe start with a comment about why US research library has such a high moat?

Steve Hasker - Thomson Reuters Corp - President, Chief Executive Officer, Director

Yes. I think -- and I think people are only just starting to understand this. So we've -- in the context of agentic AI and a deep research environment, there are two things that we think are very significant advantages. And sort of people, I think, tend to generalize and say, well what's that? It's the content.

The content that helps ground the output and correct the output of foundation models and the integration of those results. And that is true. We have the -- by some margin, the broadest and deepest repositories of content in both -- especially in Legal, but also in Tax & Accounting and other areas. The second advantage, which I think is as big, if not bigger, in an agentic environment, is a couple of thousand highly trained practice expert attorneys who help train the agents. So if you think about the first version of an agentic research product, which is Westlaw Advantage, what Mike, Dan and the team have done is they've taken a series of advanced research tasks and they've broken them down into a dozen steps, or 20 or 30 steps.

And then they've tapped into our thousands of experts to train the agent at each and every step to behave like a world-class expert. And that is not something that anyone else can do because they don't have the same depth and breadth of expertise in-house. We think it's a very difficult capability to emulate. And therefore, we think that the emergence of agentic AI has actually significantly strengthened our capability, and moved us to the right of this chart on page 8. So I think that's the first part of your question.

Mike and Gary, do you want to go to the others?

Andrew Steinerman - JPMorgan Chase & Co - Analyst

So you want me to repeat the others? Or do you want to?

Michael Eastwood - Thomson Reuters Corp - Chief Financial Officer

I think we're good.

Andrew Steinerman - JPMorgan Chase & Co - Analyst

Okay. So yes, go ahead. So the other thing I wanted to jump in on is on workflows. Like where do you feel like, given your foundation of the Legal research library, where do you feel like CoCounsel has the best ability to win within these new white spaces? And then, of course, against these new white spaces, there are new start-up competitors?

Steve Hasker - Thomson Reuters Corp - President, Chief Executive Officer, Director

So let's start with Legal and then we'll talk about -- Yes. We also made a comment about Tax & Accounting. I would say, look, where the stakes are really high and the results need to be very accurate, and/or the agent needs to behave like a world-class expert. I think we have, as I said, an advantage there, firstly. Secondly, where that workflow involves a component of research, I think we have an advantage there because some of the more generic tools that aren't trained in our content, that don't benefit from access to the experts, just don't produce accurate enough research results and many customers have sort of said, look, you can use one of these tools to their associates, but you can't use it for research.

And then I think in terms of both litigation and transactions activity, whether it's drafting, whether it's development of an initial briefing, or rebuttal, a likely reaction of a judge, the sort of workflows around specific M&A activity. I think we're well placed to compete. And where we aren't, or we won't invest in those areas, that's where our -- our dry powder in terms of the financial capacity will be helpful because we can go out and potentially acquire some point solutions to fill in some of those gaps. But Mike and Gary, what would you add before I talk about Tax & Accounting?

Michael Eastwood - Thomson Reuters Corp - Chief Financial Officer

Yes. Just in regards to the corporate segment, Steve talked about the law firms. We're seeing very similar uses of CoCounsel Legal within the general counsel space. Steve will expand more, but also within the corporate segment, we're seeing CoCounsel for tax being used as a supplement or alternative to Checkpoint. Just as a reminder, we have Westlaw, which is used for Legal research.

We have Checkpoint, which is used for traditional tax research. Checkpoint for Tax helps customers in that traditional tax research that they've previously used Checkpoint. So we're seeing customers migrate at a pretty good pace from Checkpoint to CoCounsel for Tax.

Gary Bisbee - Thomson Reuters Corp - Head of Investor Relations

Yes. The only thing I'd add, Andrew. I think you framed it the right way, which is the AI workflow space is white space, or an incremental opportunity for us. And certainly, there's some new competitors there. We're doing well.

We're winning our fair share. And our focus, and Steve alluded to this, looking forward is leveraging our content and our expertise to deliver workflow automation solutions that are difficult for competitors that lack those two assets to replicate.

Andrew Steinerman - JPMorgan Chase & Co - Analyst

Gary, have you -- you said winning our fair share and things like intelligent draft. Like have you ever seen a market sizing on intelligent draft? Like how do you know you're winning your fair share? You're growing quickly, but how do you know it's, quote, Thomson's fair share?

Gary Bisbee - Thomson Reuters Corp - Head of Investor Relations

I think we've got some sense based on how competitors have discussed it and certainly, discussions with our customers of how we're doing, and we think that CoCounsel is competing very well and is one of the market leaders. Our goal is to continue to drive that success.

Andrew Steinerman - JPMorgan Chase & Co - Analyst

Steve, I want to get back to tax research.

Steve Hasker - Thomson Reuters Corp - President, Chief Executive Officer, Director

Yes. So if you think about sort of where to where -- to Mike and Gary's points, where the competitive dynamics haven't changed in this AI era, in the Tax & Accounting space is around the calculation engines. So we have one. CCH have one and it's clearly -- it's a very, very stable space. What we have not seen is new entrants into that space.

What we have seen is the emergence of agents to start to perform some of the tasks that are ancillary to the core tax calculation. So research around changing laws and regulations to Mike's example. And so where we're going here is to use agents to perform tasks end to end, such that the first draft of a tax return, can and will be performed without human interaction. And the reason this is important because of the acute talent shortage in the Tax & Accounting professions. And so if you think about collecting all of the source input information, and making sure that it's accurate, that it's curated, that it's input into that tax calculation engine.

And that once the calculation engine is run, that any potential issues have flagged and highlighted for follow-up by the Tax & Accounting professional. And then the e-filing process and so on and so forth. So we're basically using the tax calculation engine to do the math as it has for many, many years, and we're using agents to provide a seamless experience that doesn't require human intervention. So this is another example, similar to Legal, where the core franchise -- the dynamics around the core franchise remain very attractive, and we're using AI to expand the role we play in the daily lives of our customers. And we're excited about the growth prospects, both in our Tax & Accounting business, that serves the Tax & Accounting firms, but also within corporate, which, of course, serves the heads of tax within larger corporations.

Andrew Steinerman - JPMorgan Chase & Co - Analyst

So I think you're going to say you just answered this question, but let me give it a try. I started out as a tax accountant, as you might remember, I know you started accounting also. My question is, when you look at the tax engine, you look at the tax research library you have, do you feel like the mode is as fortified as in your Westlaw business?

Steve Hasker - Thomson Reuters Corp - President, Chief Executive Officer, Director

Yes, I do because I think those tax calculation engines are very difficult to replicate. You have years of historic data. You have a profession that at least every six months needs to be running those engines and filing. And therefore the window to swap out, one of those tax calculation engine, and the cost benefit of doing so is, I think, favors the incumbent providers, provided those incumbents are providing a valuable service to customers, which we've certainly worked on over the last couple of years. So I very much like the competitive positioning as a starting point of those tax calculation engines. And I think it's as strong as our starting point in Legal research.

Andrew Steinerman - JPMorgan Chase & Co - Analyst

And you're saying you haven't seen any competitors in Tax & Accounting?

Steve Hasker - Thomson Reuters Corp - President, Chief Executive Officer, Director

Not in that core tax calculation engine. What we're starting to see is some AI agents. But again, they don't benefit from the same content and expertise that we have, nor do they -- nor can they hook into the tax calculation engine. So I'm not here to denigrate them. We're trying to learn from what they're doing as I'm sure they are from us. But we like our positioning as this space evolves and the AI agents become more prevalent in the day of a Tax & Accounting provisional.

Gary Bisbee - Thomson Reuters Corp - Head of Investor Relations

Andrew, if I could just add. If you look at our strategy pre generative AI. We've always been in the center with this really embedded position with the tax calculation engine. What we had done over the prior five, six, seven, eight years was acquired, and occasionally organically, add more capabilities to really be able to offer truly end-to-end workflow and tax. What we're doing with AI is now, if this is your workflow, choose parts of it and automate those parts of it to help drive more efficient process for our customers.

And I think nobody's got that end-to-end software workflow as a starting point. And as Steve said, we think the tax engine has a very strong moat. And to some extent, it gets larger when you've also got all these other solutions that you're selling into them. And there's a big opportunity to bring AI to automate parts of that workflow in tax and also an audit.

Andrew Steinerman - JPMorgan Chase & Co - Analyst

Great. Really liked the Legal growth in the third quarter. Is there anything that makes you feel like it would be hard to accelerate the Legal segment growth from the recent levels?

Steve Hasker - Thomson Reuters Corp - President, Chief Executive Officer, Director

I think there's sort of a couple of considerations. One is the -- we're starting to see growth in the TAM. So we're starting to see law firms and general counsels spend more on technology. In the case of law firms, I think it's more on technology and less on real estate. That's the thesis we've had for a while, and I think we're now starting to see that materialize.

So that's a factor as to how quickly that changes. And I think it will change. The speed of that change will be dictated by how quickly they can address the revenue model, the billing -- per hour billing. How quickly they're going to -- they can change the pyramid structure and the apprenticeship model within the firm. But that's a pretty exciting, I think, tailwind for us.

I think the second is our rate of product -- to your first question, our rate of product innovation and execution. And so we're growing in confidence and optimism on the back of our significant releases in August and the future -- the road maps there. And I think the third consideration here is our government business, which we report as part of Legal. And we talked about at the end of the -- at the end of September, so the year-end for the federal government. We had a couple of downgrades within that.

Our government team, I think, has done a very good job of sort of maintaining its must-have status within the key agencies that we provide tools to, but we did suffer a few downgrades. We're addressing those. Our product set is broadly aligned with driving efficiency, with preventing fraud, and with sort of law enforcement and broader security. And so there could, and should be, some growth opportunities for us. But I think as we all know, that's the unknown, and we flagged that on our November 4 earnings call for that reason.

Michael Eastwood - Thomson Reuters Corp - Chief Financial Officer

Yes. Andrew, we delivered 9% organic growth in Q3. I stated on the earnings call. We forecast 9% for Q4, absent any additional downgrades within the US federal government.

As we sit here today, we are comfortable with the 9% for Q4. We stated on the earnings call, that the US federal government had about a 20-basis point-impact in calendar year '25. We would expect that to continue into calendar year 2026, just given the nature of those subscription-based contracts, meaning a portion of the revenue recognition happened in '25, the residual will happen into 2026. As we go into the February earnings call, we'll provide additional segregation within Legal professionals of the government business. The state and local portion of government has performed well in 2025, and we expect that to continue.

Andrew Steinerman - JPMorgan Chase & Co - Analyst

I think you're saying steady, high growth in Legal, when I interpret what you were saying?

Michael Eastwood - Thomson Reuters Corp - Chief Financial Officer

Yes.

Andrew Steinerman - JPMorgan Chase & Co - Analyst

Okay. Okay. That's great. Talk a little bit about law firms. Obviously, to embrace your products, you require a lot of change, really kind of cultural change in law firms. Are you seeing maybe small and medium-sized law firms address these types of changes, these efficiency changes, maybe fewer para legals, than large law firms? Or are large law firms as forward leaning on adopting AI tools that would require workflow changes?

Steve Hasker - Thomson Reuters Corp - President, Chief Executive Officer, Director

Yes. I mean I think what's been interesting about this era of AI-driven transformation is, for us, has been, that for a product like Westlaw Advantage, in the previous versions that I've been around for, so Edge and Precision, we saw the very largest firms adopt first, and then that sort of move its way down into the mid-tier firms, and even some of the smaller firms. What we're seeing with Westlaw Advantage and CoCounsel, is the largest firms for sure, because they have the sophistication of the budget and they want to be at the cutting edge, we're also seeing sole practitioners purchase these products. And that's a new dynamic for us. And it suggests that the sort of adoption curves ought to be good news for us as they

-- as we roll forward.

I think broadly speaking, the profession has shown an open-mindedness to adopting AI tools that is greater than I would have expected. When, for example, we were at our Investor Day a couple of years ago. So I think that's actually been more than -- I mean, it's very hard to find the managing partner of a law firm who hasn't accepted and started to embrace the use of AI within their practices. Very, very hard to find. I think though, the change management beneath that is yet to really take hold.

And we're doing -- we have a program with a number of our largest and most innovative customers to really partner with them and drive that adoption and see the benefits flow through throughout the firm, including to equity partner profits. But it's one thing to license the tools and to ask associates to use them. It's another then to rewrite the way in which the practice of law is conducted. And I think that will take some time, and that's the process that we started with these luminary customers.

Andrew Steinerman - JPMorgan Chase & Co - Analyst

Okay. Great. When I look at the 2026 targets, Mike, for Tax & Accounting of 11% to 13% organic revenue growth, that includes acceleration. So my question is what's driving that acceleration? Is it price? Is it product? What's driving it?

Michael Eastwood - Thomson Reuters Corp - Chief Financial Officer

Just as a reminder, price within Tax & Accounting professional has averaged about 5% to 5.5% in the last few years. We expect that to continue as we move forward into 2026. Second, we have the recent acquisitions of SafeSend, Materia and Additive within Tax & Accounting professionals that will continue to scale and accelerate into 2026. Third, an item that has been a consistent contributor within Tax & Accounting, it's the Dominio business within Brazil. For the last 11 years, it's grown 20% CAGR. We expect that to continue. And then lastly, we expect the retention for Tax & Accounting professional to continue to increase. I think the convergence of those four items give us confidence with that 11% to 13% for 2026.

Gary Bisbee - Thomson Reuters Corp - Head of Investor Relations

I'd just add one other, Mike, which is the organic GenAI launches that we've done of ready to advise in August and ready to review, which is targeted around year-end, along with the scaling of CoCounsel for Tax, which is powered by Materia. So that was incorporated to Mike's M&A comment, but I think those GenAI launches and products we're very excited about.

Andrew Steinerman - JPMorgan Chase & Co - Analyst

Yes. No. Ready to review is definitely one. Okay. Okay. Great. Talk a little bit more about the corporate side. I know you guys keep on telling us it's the bigger opportunity for penetration than law firms. And when I hear that, kind of bigger opportunity, I kind of think like, oh, maybe it is the bigger opportunity kind of theoretically, but like historically, the penetration has been lower then? And what's the catalyst for change?

Steve Hasker - Thomson Reuters Corp - President, Chief Executive Officer, Director

Yes. So we have, as you say Andrew, for some time, said this is our biggest TAM. The opportunity to serve the Head of Tax, the General Counsel, the Head of Risk and compliance, internal audit within corporations, large, medium and small. And as you said, our starting point is the most modest in terms of our penetration. So we have conviction around that.

That conviction was strengthened last year in 2024 when we had double-digit growth. We decided this time last year to make a series of changes to go to market because we've seen 10% growth. So I think you've had a sense of optimism. We'd also seen NPS, as I said earlier, improved meaningfully within all of our segments, including corporates. What we hadn't seen was a real meaningful uptick in cross-selling.

So we made some changes to try to improve that. In doing so, we probably created a little bit more disruption than we'd like, but we're starting to see the green shoots of those efforts. We're starting to see that book of business build in ways we like. So look, I remain optimistic that it is our biggest opportunity, and it won't be one of those ones that will always be our biggest opportunity. I think we'll start to see that growth fulfilled. The other thing we --

Andrew Steinerman - JPMorgan Chase & Co - Analyst

What's the catalyst? Is it -- you have better products for them? Like what's the catalyst for change?