William Blair Thinking Presents—Artificial Intelligence in Information Services: Evaluating the Ramifications of a New AI Paradigm

In this podcast episode of William Blair Thinking Presents, global services analyst Andrew Nicholas examines the industrywide impacts, implications, and potential risks that artificial intelligence presents for the information services sector and offers an overview of the framework he employs to evaluate information services companies in light of these industrywide dynamics.

Podcast Transcript

Chris Thonis: Welcome to William Blair Thinking Presents new podcast series that aims to provide in-depth expertise from our award-winning equity research and capital advisory teams on today's financial and economic landscape. I'm Chris Thonis, head of equities marketing and media relations, and I'm delighted to be your host. Hello, everybody. On today's episode of William Blair Thinking Presents, we welcome Andrew Nicholas.

He's our equity research analyst on the global services team. He covers consulting, HR technology, and information services companies. You know, as everyone may recall a little while back, we hosted analysts Ralph Shackart, Jason Ader, and Arjun Bhatia. They were on the show to detail the report on generative AI and its implications for the tech sector. And as a follow up to that, I thought it would be fun to get Andrew and his team, at least Andrew right now, to help walk through their new report, which addresses how AI and generative AI will impact the information services industry and to establish a long-term framework

for evaluating an information services business and its underlying data assets in this new AI paradigm. So, Andrew, welcome to the show. To kick this off, do you mind explaining first and foremost what the information services sector is? I think that would be helpful and then from there providing a brief overview of what this report is all about?

[01:29]  Andrew Nicholas: I'd be happy to, and thanks, Chris, for having me. We'll start with describing what an information services company is. I will admit it is a pretty broad scope. Firms that provide data analysis, research all fall within the information services bucket. Usually that data and that analysis is going to some sort of decision maker.

But the type of end market or the vertical or the industry is really broad and it just depends on the company. Obviously, this is mostly a B2B business or a B2B industry, but you can get data and analysis and research out to the consumer as well. But historically, this has been a B2B business and that's the sector that we cover here at William Blair.

In terms of what the report is and why now, I think both of those questions are important. And I'll start with the why now. I think increasingly, as the technology has become more and more ubiquitous within kind of the broader day to day life of individuals across the country and across the world, I think chatGPT and similar tools has really increased how familiar people are with artificial intelligence and the technology, and it's made it a lot more interactive.

People are able to engage with it in a way that people without, very advanced degrees and capabilities were only able to before. And so as that has become a mainstream topic, it became an important investor question as well. And with information services and data and its leverage of artificial intelligence already, it made it a natural extension for us, or at least in my opinion, to cover exactly how AI and generative AI will impact the information services industry and then also the leading companies within it.

And as I think we'll talk about as we move through here we came up with a bunch of different factors that we think are most relevant in answering the question, “Will this company benefit from artificial intelligence? Will this company benefit from generative AI, or is it at risk and in some way, shape or form.”

[03:44] Chris T: That’s a great segway. So there's three primary impact areas you tend to focus on here with regards to the impact of AI on information services. Those include increased data content, product supply, increased accessibility, and then meaningful improvements to operational efficiency. Do you mind diving into those a bit deeper for the listeners?

[04:07] Andrew N: Yeah, absolutely. I think as you as you outlined, there's three bigger tailwinds here as far as we can tell. Not only do we think AI and generative AI increases the supply, it also increases the demand for that information. And then lastly, there's the third chunk, which is operational efficiency. I'll hit each of those in order here.

First on the supply side, so there already exists data information out there in the world. The companies that operate within this industry have made a business in collecting that information. And collecting that information is core to building the data assets that make them so valuable. I think artificial intelligence has the ability to streamline the collection of that data, streamline data aggregation, and in doing so, can really improve the value and the scope of the data products and the data assets that these companies are compiling.

And before even getting into the other pieces of the increased supply conversation, I think it's also important to recognize and I should have mentioned this earlier, information services companies are not new to artificial intelligence. This is an industry that has leveraged machine learning and natural language processing and AI to build their assets, clean their assets, link data for some time.

But it does feel like this has reached a tipping point, at least in terms of the investor lexicon. I wanted to double back there first.

And so, you have this data collection capability that's important to building these information services companies assets. Now, in addition to that, you have generative AI allowing for content generation. And companies within the space that produce research reports, that produce information, summaries that analyze data and provide it to those same end users we talked about earlier are now able to use this technology to do so.

And I think as we'll talk about later there, there are efficiency gains to this. But more than anything, the more content is better in that they can sell more of it to those same customers or even potentially create content that they wouldn't have otherwise created, which could open them up to new markets, new users and new consumers of that information.

And then the last piece of the increased supply equation that we touch on is, I think, artificial intelligence in general helps with the identification of patterns in data. And as artificial intelligence gets more sophisticated, perhaps enhanced by generative AI, I think that companies within this ecosystem will be able to create new products and more than anything, they can fail faster with product development, being able to figure out what is going wrong and figure out a different way to approach it.

[07:03] Chris T: How are these just out of curiosity, how are these patterns being discovered in the past? Was that just done simply through hoards of data processing by the people at these companies, or how was that done prior to machine learning or AI?

[07:18] Andrew N: So yes, I think this has been such a critical part product development for information services companies for some time over a decade that I'm not sure that I have a ton of context in terms of that question. But I would say that that answer the answer to it is very similar to what I mentioned earlier, which is these are companies that live breathe data each and every day.

And so, they've been at the forefront of data analysis and data aggregation from the get-go. And that forefront just seems to be moving further and further ahead. You have the first dynamic, which is you have increased supply. The more data, the more that you can sell, the more content, the more that you can sell. But then there's this other piece of the equation that is this increase in demand.

And I think this is really the unique opportunity that generative AI specifically creates. I think that generative AI makes information more accessible, whether it's improving the search and discovery functionality, whether it's improving how easy a product is to use. I think the big case here is that a user that is not a data scientist, that does not have a credible level of acumen in that area.

They're now able to access the same high-level, detailed in many cases, proprietary information, and that expands that user base. And when the accessibility increases, I think you increase usage, I think you increase retention. And that's ultimately going to be good for demand for information services companies. And then the other piece of the demand equation is just the democratization of artificial intelligence broadly.

I think that there's lower barriers to entry for using artificial intelligence now. I think that customers of all sizes, a bunch of different industries now have the ability to access technologies that would have otherwise been cost prohibitive or required specialized expertise. If you take increased supply, particularly with data and content and product that's hard to replicate and then you increase demand for that same information, I think it really has the potential to meaningfully enhance the value of both the individual data assets that these companies own, and also the companies themselves.

And that's the demand supply equation. And then on the operational side I think software development, customer service, sales and marketing efforts, all of those things can be augmented by generative AI. And I think that's something that enterprises across industries, even outside of the information services space will benefit from as well.

[9:59] Chris T: I think it's hard to have a conversation about AI without talking about the inherent risks, right? That exist for the sector. So why don't we tackle that real quick? In the report, you outline three primary risks to the information services sector. Its AI specific risks increase potential for competitive disruption and then resource constraints.

We'd love to know what are these risks in particular? Just maybe if you just detail each of those and then how can these information services companies address them both in that near-term and long term?

[10:32] Andrew N: Yeah, that's really important. Obviously, we're pretty excited about the technology's capabilities and its impact on the space, but there are some pretty important things to consider in terms of risks and costs. So again, I like to think in threes, and I think there's three primary ways.

So, the first is that it dovetails off the supply conversation that we had earlier. Content generation’s easier, data collection might be easier, new product to innovation improves. I think the first risk for information companies or information services companies, particularly the incumbents, is that it could spur innovation outside of those companies. It might be easier for firms to generate or create data assets and data products than it was previously.

And with that comes increased competition. And increased competition, the potential for competitive disruption is something that all investors need to be cognizant of as they look at companies within the space. I would say in terms of addressing that issue, I think these data assets and these companies have always been very protective of their competitive moat in a lot of the things that would help them limit the amount of competitive disruption are something that we cover in our evaluation framework, which I think we'll talk about later.

So that's the first thing is competitive disruption. The second is AI-specific risks. And I think these are the ones that you typically hear about in the news or get the headlines, things around data privacy and security, hallucinations, basically AI models making things up and not providing correct information or perhaps misleading you.

[12:22] Chris T:  I like that you call it hallucinations. Is that the industry term?

[12:26] Andrew N: I think that's one that that also our colleagues in the tech group leveraged in their report and this referenced previously. But yeah, it's basically a generative AI model or some sort of large language model gives you an answer could very well be making it up.

[12:40] Chris T: If it's so, that’s a little scary. But I have heard that yeah.

[12:45] Andrew N: And in an information services company which its brand and its value proposition to clients is saying, what we're going to give you is highly valuable but it's also accurate. Hallucinations would, would certainly be a threat to that. And then there's also biases within that that these firms need to be cognizant of. I think that their solution or addressing those issues ultimately comes down to the information services companies focusing on data security, focusing on data accuracy, focusing on explainable AI technology, sourcing their information.

All these things are helpful in eliminating some of these risks and are things that the biggest information services companies are already very much focused on addressing. And then the last piece is just resource constraints. This is it's a little bit outside of my area of expertise. I'm sure there's technology and infrastructure analysts on our team that would be able to speak to this a little bit more eloquently.

But I think you have to be cognizant of the fact that leveraging these models a lot of times using or paying cloud providers for the compute to be able to process large data sets and do some of the things that we've talked about already doesn't come without cost. It's pretty expensive. And so that's one consideration. I don't know if it's as much of a risk as it is something to consider because information services companies are going to have to kind of balance the cost benefit of leveraging large language models in the future, particularly if supply on the chip side or the compute side is limited.

[14:23] Chris T: Makes sense. So there's a lot more to this report. Obviously, there's a whole in-depth analysis that looks at how I may impact ten of the information services companies that you and your team cover. But we don't have time to dig into that, nor can we. We generally can't talk about companies specifically, but it's probably worth running through the evaluation framework that's laid out in the report around data assets and then the company and then companies operating within the paradigm.

If you don't mind, if you just provide some key takeaways there and just reword the question as the evaluation framework laid out in the report, that is around data assets and in companies operating within this AI paradigm.

[15:05] Andrew N: Yeah, I think this is something that we're really excited about in terms of the report. We've covered the tailwinds to the sector, we've covered the risks and additional considerations. I think that leaves us with the question, who will benefit the most and who is most at risk? And I think in trying to answer that question, we found the best way to do it was to come up with an evaluation framework for information services companies.

I think in doing that or performing that exercise, we basically learned two things. First, that a lot of the factors that determine a company's success within this new AI paradigm are very much applicable to how we evaluate information services companies already. To say that another way the evaluation framework overlaps with an analysis of a firm's competitive moat in a lot of ways.

And then the second learning is, is the factors that we think are most important to answering this question. And as you alluded to, we came up with eight factors which we split into two buckets. The first bucket a four is focused on the variables to consider when evaluating the quality of the data asset itself. So those factors are how proprietary the information is, basically how exclusive that information is, how difficult it would be to replicate that data set, how accurate it is, which is, how reliable the information is.

Free from errors, free from biases, which we talked a little bit about earlier, how rich the data set is, or data asset is the breadth, the depth, the timeliness and the completeness of the information. And then lastly, the applicability basically to say how many different use cases are there for this information? How many different end markets or industries could find value in the information that is contained within this data set?

Ultimately, those first four variables will determine how valuable the data asset is and the likelihood of replication over time, which, again, is very important when we're considering competitive disruption and potential benefits in an AI paradigm or this new AI paradigm. And so that's the first for that, that's the quality of the data assets. But then there's this other segment of our evaluation framework, which refers to what matters the most at the product or the company level.

These are the variables an organization can bring to the table to either help extend or enhance data assets, competitive positioning. So again, we're going to four variables, scale, the size of the company, basically, brand, which is, as you well know, the perception and reputation of the company in the market, the stickiness of the products that they have which is basically to say, how easy is it for you to switch off the product that you're using, how mission critical is what we're providing you as an information services company to your day to day operations?

And then lastly, infrastructure. I think a company's data and technology infrastructure plays a large role in how well a company can collect, store, and really manipulate large volumes of information both efficiently and securely. So you take all these factors, there's eight factors and we have the whole evaluation framework in our report and we go through all the different companies we cover and evaluate them on these metrics.

But in the end, I think if you score well on these eight factors, you're going to be positioned for success in this AI new paradigm. And that's really the genesis of the report and what we set out to do.

[18:52] Chris T: Before you go, unfortunately we are down to our last minutes. I think it'd be helpful if you could take everything we just chatted about and then maybe just provide a synopsis of sorts that synthesizes the tailwinds, discussion, the evaluation framework and then put it into takeaways for the full sector.

[19:10] Andrew N: Yes, I'd be happy to. I think there's a few points to make. First, it's generative I think is a game changer. It's not instantaneous and I think the use cases for Gen AI will evolve, but I do think it adds fuel to the AI fire, if you will, and I think a lot of the trends that were in the space already are going to be accelerating.

We talked about the three major tailwinds, increased supply, increased demand, and improved operating efficiency. I think that's the first takeaway. I think the second is that proprietary data is important, but it's not the only factor. We went through all eight of the factors. I think it's important for investors to keep in mind that while having something that other people don't have is kind of step one, there are other ways to provide customers with value beyond that factor.

And then the last thing is just how this impacts the entire factor. I think to kind of sum it all up, I think there's faster revenue growth that can result from this. You have new products, you have product enhancements, you have better customers who are able to access the information more readily and you have increased value of great data which could potentially result in some sort of pricing opportunity.

So you have faster revenue growth plus the operational efficiencies that we touched on to expand margins. And I think ultimately over a multi-year period that can result in faster earnings growth for the group as a whole and particularly for companies that score well on our evaluation framework, which we detailed.

[20:48] Chris T: Well, Andrew, always great to chat with you. Let’s do it again soon. We'd love to have you back. But beyond that, thank you again.

[20:55] Andrew N: Thanks, Chris.

[20:57] Chris T: Yeah, you've got it. We'll talk soon.

For more head WilliamBlair.com/thinking where you can browse our library of white papers, market updates, webinars and all these other resources designed to provide actionable intelligence for emerging opportunities. If you like what you heard, share and subscribe wherever you get your podcast. Copyright 2023 William Blair and Company, L.L.C.

William Blair and Rdocs registered trademarks of William Blair and Company LLC. As used on this podcast William Blair refers to William Blair and Company LLC. William Blair Investment Management, L.L.C and Affiliates. For more information about William Blair, go to www.WilliamBlair.com. This content is for informational and educational purposes only and not intended as investment advice or recommendation to buy or sell any security. Investment advice and recommendations can be provided only after careful consideration of an investor's objectives, guidelines and restrictions. The views and opinions expressed are those of the speakers and are subject to change over time as market and other factors evolve.