Transcription
Hi, I'm Matt Murphy. I'm senior vice president of client advocacy at Ibisworld. And today I'm joined by Andy Brennan. Andy is chief data officer here at Ibisworld. And today we're going to be talking about API. Uh, because there's a big problem out there. And that problem is that organizations these days have a lot of data at their disposal. They have data about their own operations and finances. They have data about their customers and the different markets and industries that those customers operate in. And they're trying to harness this data in a way that puts that information in the hands of the right people within their organization, uh, at the right time so that they can make more informed business decisions. So, a lot going on in this space. Um, I, Andy, am somebody that, you know, likes to get, uh, you know, the bad news before the good news. So, I'm actually going to start us off with some of the downside potential. Uh, as a chief data officer yourself, do you mind talking a little bit first about what some of the risks there are inherent in trying to integrate this type of data with API?
>> Sure, Matt. Um, great to chat today. The biggest consequence of bad data is bad decisions. Once you have bad data in your commercial systems, it can flow right through the chain, can distort forecasts, can trigger the wrong actions, and ultimately create errors for every team that uses the data. So, especially with, you know, this new world of AI, as we all know, for anyone that's that's used AI, trust is a big factor. AI can sound confident but often, uh, the information it's providing is is just flat wrong, can hallucinate. So, yeah, the biggest issue with, uh, with bad data is bad decisions.
>> Yeah, and I, I think that can come in a few forms, right? The the data itself could come from an untrustworthy source. Um, we know that a lot of these AI tools and these models are pulling from a variety of sources and it's not always clear what the original source of the data was, or the data could simply be outdated, right? You know, you could be, uh, using a statistic that is from 3 years ago when you are essentially, you know, assuming that it's, you know, data that's relevant today. As far as IBIS world goes, Andy, um, you know, a lot about our process and the data that we use. Can you talk a little bit about how IBISWorld ensures credibility and trustworthiness of of its own data? What is our process for for verifying our own industry data?
>> We've spent the last 50 years doing one thing, which is turning messy, inconsistent data into industry intelligence. So, going way back before the internet, you know, when when we used to go and grab books from the Australian Bureau of Statistics back in the in the 80s and 90s, bring them back to the to the office and plug them into our our database. We're still doing that today, but obviously with a different process behind it. What that kind of looks like, Matt, to give you an intro into the the secret sauce at Ibisworld, we've got nearly 100 industry analysts. Their full job is to model, interpret, uh, industry trends. They then bring context, uh, judgment, and domain expertise to our data so that they're writing, um, up-to-date analysis for our clients. So, on top of that, Matt, we've also got a data engineering and analytics team, uh, whose main job is to bring in all the different sources we use. We use up to 400 sources. Everything from government agencies to paid sources from our partners through to proprietary data. We bring all that into our data vault that gets cleaned up, organized, made consistent, uh, mapped to our unique industry taxonomy, and there for our analysts and our clients to use. So that that's effectively the secret sauce. We have our our data engineering, our analysts on top of that, and then our QA and editorial team that really make sure the data is accurate and fact-checked.
>> Yeah. And internally, we're we're ourselves using API to to collect a lot of this data and and put it in the hands of those, you know, real live human analysts that you mentioned, of which we have about a hundred or so team members. Um, so, so let's say I'm an organization and and I'm very confident in my internal data. I'm very confident in the third-party data like industry data from IBISWorld. Um, and I want to, you know, take that data and I want to make sure that the relevant pieces of information are getting to the relevant teams or team members. Um, API is a really powerful way to do that. So, so what is API and what is the process with which companies are building APIs into their data workflows?
>> Yeah, good question. API is really the backbone of the internet. So, every time you check the weather or order an Uber, really, it's an API that's talking to another system. So, APIs are really just messengers that different software systems can talk to each other. So, as you mentioned, Matt, we use APIs a lot here to gather data sources. Uh, we, um, our systems talk to whether it's a government agency or one of our partners' systems to pull the data into our system, and it hides all the complexity so that we don't need to know, uh, how the other system works. Uh, we just, uh, hit hit their API and query their data straight to our database. So, I always like to say that the API is sort of like a waiter in a restaurant. You don't need to know what's happening in the kitchen. You, the customer, talks to the waiter, and the chef prepares. The waiter brings it back. So, you don't need to know all the hustle and bustle or how the how the chef's cooking that meal in the kitchen. You just need to know what's on the menu.
>> Yeah. Yeah. And I think what the major, you know, um, you know, application for this is is simply just, you know, saving your employees time at the end of the day, right? Because, you know, if I'm a researcher and I've got a project, you know, I might have to go to, you know, 11 or 12 different data sources, sometimes more in those cases. I'm logging into different resources that my company has provided to me. Whereas if we have API deployed and we know that there are certain, um, data sets that we want to access for these types of projects, that API can very easily be called in, and next thing you know, rather than having to log into all those different resources or go hunting through a bunch of different data sets, all the information that I know that I'm going to need is delivered right to me. Whether that's in a dashboard or that's in an internal resource that a company makes available. Um, those APIs could could even be called into an AI tool, for instance, which has certainly become a very popular way for researchers to to query the data that their company makes available to them. Um, Andrew, there are there specific use cases that you've seen IBISWorld clients using API for?
>> That's a great question and we've had an API for more than 10 years. We first developed it back, uh, about 10 years ago, uh, when we worked with a big bank to automate some of their credit decisions, and it's only grown since then. We've now got hundreds of clients using our API. There's three main use cases at the moment. Uh, one is to automate decisions. So, things like credit decisions, loan approvals, where banks will pull our API into their internal systems. They often are using our industry risk, uh, data or any of our industry data that plays a part in credit decisions. So, that's number one is to automate decisions. Number two is to enrich dashboards, uh, and BI tools. So, often our clients need to understand the industry growth as part of any of their internal dashboards. That's another main use case. And increasingly, the third use case is to power, uh, AI tools. So, clients need to ensure there's like a verified, validated, uh, industry component to their, uh, AI tools. Often they're pulling in lots of different data sources from other vendors, and IBISWorld's that industry piece that we mentioned before. AI is a really popular way for clients to leverage our data, and as you mentioned, it it really speeds things up. There's no coming to our website, uh, individually and analyzing an industry. You can have it right there in your system.
It's so funny that you mentioned, um, you know, 10 years ago when we worked with that bank and we we provided our information via API for the first time. You and I have both been at IBISWorld for a dozen plus years. I remember sitting in a room and people describing this thing API to me and thinking, you know, what is this? You know, this is probably something that's only going to be used by, you know, the biggest and most global of organizations. But I think what's really amazing and what's happened over the last 10 years is, you know, we have small organizations that maybe only have five employees that are, you know, pulling data sources like Ibisworld in via API and being really nimble with the way that they're digesting data into AI tools. So, it's really been a journey that the market's been on for the past several years, a journey that Ibisworld has been on for the past several years. And, um, you know, banks now, you don't need to be a big global bank. You know, we have small community banks that are, um, taking our risk scores and baking them into their own risk models using API. You know, they're not having to lift a finger. And Ibisworld's, uh, monthly updated risk scores are embedded right into their risk models, and they're able to make, you know, more informed lending decisions and portfolio decisions based on having that industry data delivered right to them. You know, we have software companies and manufacturers that are trying to, you know, understand what markets they want to tap into. And IBISWorld's, you know, industry market size data is being fed into these dashboards that sit right alongside information about their customers, which I think is really exciting. You'll be looking at an individual company's information and, you know, right next to it could be IBISWorld's industry call prep questions. So I've got all this information about this company that I want to call on, and right next to it is, you know, these three or four great questions that I could ask somebody at that company about what's going on in their industry. Um, I think also benchmarking has been a big thing for us here at Ibisworld. You know, um, we have accounting firms and banks that are, you know, constantly looking at the financials and balance sheets of, uh, their own customers. And being able to have IBISWorld's industry averages right next to that, um, is a great way to say, okay, clearly this is a company that is underperforming its peers in the industry because I can see them right side by side, or maybe this is an industry, or sorry, a company that's outperforming, uh, its peers in that industry. So, um, it's not one, you know, industry that that's using APIs. It's not just big companies that's using APIs. It seems to be just about every type of industry and every type of business size these days.
>> Yeah, that's that's a really good point. It's not just Fortune 500 companies that can use our API to to automate their decisions and plug it straight into their systems. Our API is really easy to use. You get your username and password, go straight to the website, and you can pretty much start using it immediately. Every organization, every size. It's a great fit if if you're trying to really scale the way you use industry intelligence.
>> Yeah. And it's a it's a great process, you know, to go through because it is so simple. You know, at the end of the day, you know, you simply reach out to your IBISWorld client relationship manager. They can put you in touch with our solutions team, who is happy to kind of uncover, you know, what industry data you need to pull via API and make sure that those, uh, data sets are being called into the the right places. Andy, if you could leave the audience with with just one message regarding API, what would it be?
>> Look, I think the companies that win going forward, not the ones with the most AI. They're going to be the ones with the the best data powering that AI. Again, Ibisworld's really easy to plug into any sort of AI system that you're building to ensure that you've got verified industry intelligence. That's that's simply it. I I don't think that a lot of our clients are building, um, lots of fancy, uh, AI systems at the moment, and they're really leaning on us to make sure that there's no hallucinations and and the data that that comes back is is accurate. So, work with your vendors like Ibisworld to in, and we're here to help you build those systems to to make sure that the data coming through is accurate. Credible data is first and foremost, and from there, it's thinking about that delivery method, of which API can be a big piece, especially if you're working with several different data sets.
So, Andy, I really appreciate you talking with me today about API. Um, it's certainly one of those things that we're seeing more and more IBISWorld clients, um, start to implement. So, um, you know, providing them with, you know, a little more clarity as to the benefits of API and and how much time it can save their team members and how it can make sure that they're using, you know, more credible and trustworthy data, I think goes a long way. So, um, I appreciate your time and, uh, look forward to talking more in the future.
>> Thanks, Matt. Thanks for the time.