#310 The State of BI in 2025 | Howard Dresner, Godfather of BI

When I talk about business intelligence, a term that I coined back in 1989, it was always intended to be an umbrella term. So it's not tied to a specific technology. It's really about fact based analysis which is driven by data. We've seen that evolve a lot over the past few years. And now of course, you know, you look at things like agentic analytics and conversational analytics. Well, that's all part of business intelligence because they the intent is to deliver the same capabilities. How do we deliver data driven fact based analysis? We have so many external forces operating our organizations. Data is your friend. You want to know when things are changing and how they're changing as quickly as possible so that you can react. I mean, at the end of the day, that's the whole point of business intelligence. How do we respond to change and execute as quickly as possible?

Welcome to Data Framed, this is Richie. Business intelligence powers every business from startups through to enterprises. So sometimes it's helpful to know what other organizations are doing with bi. I'd also like to know whether BI products are getting better, whether we're making any progress towards self service analytics and whether newer related technologies like semantic layers are making a difference. The man with the answers is Howard Dresdner, AKA the godfather of business intelligence. He popularized the term business intelligence back in the 1980s and back when he was a Gartner analyst. He's one of the greatest influences on BI becoming a ubiquitous activity these days. He's Chief Research Officer at Dresden Advisory Services, one of the leading creators of thought leadership on BI data and AI. Today we're going to be looking at the results of his company's recently released the Wisdom of Crowds BI report. Let's find out about the state of BI in 2025. Hi Howard, welcome to the show.

Thanks Richie. Good to be here.

Wonderful. So I want to dive straight into the juicy bits. Can you tell me from your research in the Wisdom of Crowds Report, what was the most surprising thing you found?

Well, so we've been doing this for 16 years now. I guess some of the biggest surprises, maybe they're not surprises, is the low level of penetration. After all of these years, we're still on average below 40% penetration on average within the organizations that participate in our research. And you know, I've been covering this space for almost 40 years now and I would have thought we would have done better by now.

That is actually pretty incredible. So you're basically saying that 60% of organizations don't have a business intelligence solution. That just amazing.

I would say 60% of the users within most organizations. So basically we have an information underclass within most organizations. So where are they getting their information when they're making decisions? Well, we don't know. Secondhand information or maybe they're just guessing. That's a little bit troubling. I mean it's, it's improving and there's a lot of promise out there that it's going to get better, faster. But that's, that's where we are after, you know, tracking this market for the last 16 years.

That's fascinating. I sort of assume that business intelligence is a relatively mature space and so it's more or less ubiquitous. Apparently not.

Okay, well all organizations have it. The question is within organizations who has it. And my belief is, and a term that I've used for many, many years is information democracy. I think everyone in the organization and all the stakeholders for that matter ought to have access to timely, relevant, data driven insights. But you know, we know that they don't. So it's an improving picture. It's just going a lot more slowly than we would have thought.

Okay, well, yeah, I guess I'll take improvements but I'm right there with you that it should be everyone should have access to these sort of insights.

Yeah.

All right, so maybe we'll back up a bit. Can you talk me through what the goals of your wisdom of crowds report are and who you interview for these things, who you survey?

Sure. So the objective is just to talk about state of the market, where things are, what the perceptions and intentions are, the use cases, the investments, where they find success, how they measure success. And so it's, it's pretty broad based. So it's, it's a global study and we survey all of the relevant users. So all functions, all industries, all size of organizations. Well, you know, just about, you know, know everyone that has a stake in business intelligence.

Okay, so it sounds pretty comprehensive. I'd like to know a bit more about who has a stake in business intelligence. Is it like particular departments, particular industries? Like who cares more about BI than.

Everyone else, more than me? Maybe nobody. But you know, I think that most organizations, most industries do care about it from a functional perspective. Executives are always at the top of the list because they're obviously have a great need to know what's going on across the enterprise. But I would also say finance really high on top of that list. And you know, as you work your way down, of course I think there's always strong interest, but maybe not as strong as those functions from an industry perspective, you know, all industries. But I would say financial services has always led in their, you know, interest and need and demand for things like business intelligence and analytics. And then from a geographical perspective. Well, I would say that, you know, the US used to lead, but I would say EMEA is, you know, very close behind that. And we've seen a lot of interest from Asia Pacific in recent years, so that's been picking up. Latin America tends to be a bit behind.

Okay, so that makes a lot of sense to me. I think like if finance in general is often hot on picking up on the latest technologies. Well, even though BA is maybe not quite the latest technology, but it's always sort of leading the space.

Well, let's, let's step back for a moment. You say that when I, when I talk about business intelligence, a term that I call Bitcoin back in 1989, it was always intended to be an umbrella term so it not tied to a specific technology. It's really about fact based analysis which is driven by data. And so we've seen that evolve a lot over the past few years. And now of course, you know, you look at things like agentic analytics and conversational analytics. Well, that's all part of business intelligence because the intent is to deliver the same capabilities. How do we deliver data driven fact based analysis?

Okay, so that's interesting. Maybe it's worth the sort of defining some of these software categories that are involved then. I think to me, like business intelligence is largely about the dashboarding tools. You've got your power bi tableau, look at all these sort of things. But then do spreadsheets count like the workflow tools?

Well, sadly yes. I mean, spreadsheets are always going to be there at the end of the world. There will be cockroaches and spreadsheets. Many software companies have tried to, you know, manage spreadsheets and organizations that matter manage spreadsheets out of the equation. They're not going away. We've actually done research on spreadsheets specifically. They're simply not going away. They're strategic to a lot of organizations and a good adjunct to other tools. In fact, all of us use spreadsheets every day. I mean it's, that's just a fact.

Yeah.

Is that business intelligence? Well, sure, because it does deliver on that promise of data driven fact based analysis. But you're right, the traditional tools that do dashboarding and reporting are certainly the most well known. But increasingly, as we look at things like generative AI and conversational analytics, a lot of the new Tools that have popped up that do that, as well as copilots and chatbots for traditional tools that also does fit the definition.

Yeah. Are there any of the categories then? So we've covered dashboarding tools, we've covered spreadsheets, there's some fancy new AI tools. What about the. I mean, there's a lot of like, workflow type tools for data processing. I'm thinking like Alteryx, Knime, that sort of thing. Do they fall under the category of business intelligence?

Yeah, Alteryx certainly would. And you have to look at it broadly. When you look at business intelligence, you also look at all the infrastructure that supports it that the users may never see, by the way. But it's there because. And from our strategic initiatives, we see that increasingly data is front and center. I don't think you can be successful with BI and analytics unless your data isn't reasonably good shape. In fact, a new term that we started to use that organizations have to become data athletes. They're kind of, you know, flabby and out of shape in most cases, but that's, that's part of the overall ecosystem, if you will. You have to have really good analytical data infrastructure to be successful. And I know we all get caught up on the, on the fixtures, if you will. It's like when you're building a house or renovating a house, everybody's concerned about the fixtures, the sinks and the like. And you know, but you have to focus on the plumbing. The plumbing is really, really important. If you, if you skimp on the plumbing, nothing really works. And I think, you know, we see that in organizations that their data infrastructure really isn't where it needs to be to support, you know, what we would consider hyper decisiveness. In other words, really being able to take strategic advantage of their data from an analytical perspective.

Okay, I really do like that analogy with the plumbing. Like, you don't want fancy gold taps and then have cracked lead pipes behind the walls. That's gonna be disastrous. So maybe we'll talk more about the, the infrastructure and the tooling side of things then. Have you seen progress in this area? Companies are getting better at building out that, that tech infrastructure?

Well, certainly some are, and they're doing a good job. And I think there's a lot more tooling out there to do it. And, and I think artificial intelligence can help in that regard. But I do believe that organizations, if they want to be successful, really need to take stock of the data. In fact, I've spoken to a number of groups recently within the last few months that are obviously very focused on, well, how do we take advantage of this AI revolution? And my answer to them is, well, take stock of your data first, because if your data is not in good shape, your AI won't be in good shape either. And so I think that there is a heightened awareness and we see that within our strategic initiatives too, that data quality and data integration is really important. And boy, there are so many more sources of data and so much more volume than there was even a few years ago that organizations are struggling with that. I think once again, I think there's better technology out there and there's more intent, but organizations are still struggling with how do we get the full value? First of all, how do we take stock, what are all the data sources? And then we, how do we get full value out of those data sources?

Absolutely. It's always tricky stuff is just going well, okay, we've collected some data. Now how do we turn that into some sort of business value?

Yeah, and the pipelines and the workflows that you refer to, that's all part of it. You do still need to have data engineers and people that really understand how to integrate all the sources of data and transform those sources of data and feed those to the various applications and users that will make use of them.

Absolutely. So since you mentioned AI, I'm curious as to how that's affected the, the interplay with like with business intelligence. So I'm wondering, because AI stolen so much of the hype and the attention in the last few years, does that mean companies are investing in AI at the cost of less investments in bi, or has it been a sort of rising tide situation where it's improved by situations?

Well, I would say that it's, it's a mixed bag for sure. There's tremendous interest and organizations have invested and I've spoken to organizations that have gotten tremendous value out of AI. But the mainstay of business intelligence remains the things you referred to earlier today, reporting and dashboards and visualization. And very often there's a co pilot sidecar that's added on. There are folks using it, some are certainly. But in terms of taking full advantage of AI, I think organizations are still in most cases, when we look at adoption of AI, what we talk about today, generative AI and reasoning layers. I think it's still very early days and of course you have a lot of, you know, vendors out there and other analyst organizations beating the drum around AI. But adoption is still early because organizations are really trying to understand what are the issues, how do we govern this, how do we secure our data? And so I think it's going to take a while before we see broad based adoption. But there are certainly use cases out there for generative AI applications. I mean, the traditional predictive AI that's much more well understood and much more well established and that's certainly in use. We know what the use cases are. But the new generative AI applications, that's a sort of terra incognita for many organizations.

Absolutely. But it sounds like there's a kind of the moral of the story then is you really need to figure out your data governance situation if you want to be able to take advantage of generative AI.

Well, and it's more than that, I would say this isn't magic. Right? If it seems like magic to you, you haven't done your homework. So understand how this technology works. How does generative AI really work? What is a vector database? What are nodes and edges? Where are the best use cases? What are the security issues? What are the issues with hallucinations? So really you need to educate yourself. And the challenge is in many cases this, these capabilities are coming in the door through existing applications that you've already invested in. So you need to understand whether or not you want to use those or perhaps you want to limit the use cases or limit the users associated with it. But definitely educate yourself before you embrace or adopt these things.

Absolutely, yeah. So that's interesting, talking about generative AI being kind of thrust upon you by all these software vendors and certainly like Microsoft's got a copilot and everything. Google's putting stuff in Gemini everywhere. It'll fit and same with almost everyone else in fact. So yeah, talk me through. Do you have advice on like when you should be taking advantage of these tools and when you might want to turn them off?

Well, I think if an organization has governance policies, your governance policies need to apply to generative AI as well, you know, and most organizations do. And so I think that governance is the key. How do we govern these things? How do we train these things? Who has access to them? What data do they have access to? Because you don't want to give broad based access to generative AI because then you end up having issues with PII and security issues. So I think that, and you know, we read about this every day, I certainly see it every day around potential security issues. So I think organizations need to use the approaches they've already used for governing business intelligence, for governing their data and how that data is utilized within the organization and apply that to AI. In fact, we just released today our data and analytics governance report, which does go into much of that. Governance is the key. So you know, we talked about being a data athlete. You also have to be a governance athlete, I think as well.

Absolutely. It's not like the most exciting topic, I think for many people. Like there's not many people like, yeah, I love data governance unless you work in that field. But it is incredibly important. Just have that kind of basic data governance sort of literacy.

Absolutely. And given all of that, I do think that organizations can get use from this and we talk about it from a business intelligence perspective and things like, you know, certainly conversational analytics being able to increase the penetration of usage. That's valuable because we're not going to be able to increase data literacy by much over time. I think that's something that sort of happens organically and one of the ways you can do that is giving people access to data. Well, how do we do that? If they're not going to use a dashboard or even a standard report, we can give them conversational analytics. So I think that you'll get greater usage as a result. And I believe that once again, going back to the whole concept of information democracy, if everyone has access to their relevant data, not all data, but to their relevant data, that helps align them with the mission and the strategy of the organization, which inevitably is a good thing. And same thing with giving it to customers and suppliers. That's a great way to get alignment with them as well. And I do believe that AI and generative AI can help in that regard.

That's incredibly important. The idea of aligning your business intelligence use your artificial intelligence use with business strategy and just returning to your wisdom for crowds report. I'm curious as to, are there any particular business initiatives or business goals that tend to drive BI usage?

Well, everybody says better decision making and that's sort of the, the odd one that's in there. There are things that are really measurable. So things like return on investment and especially when you're trying to, you know, save money, you have an existing process and you know what that process costs you. It can somehow decrease how long it takes you to do something. So ROI is one of those measures. And I do think that it's one of those things. Organizations are certainly focused on better decision making, but also increasing revenues and decreasing expense. Right. Those are the two, only two things that matter really in business. Right. Making money, saving money. And those two initiatives tend to go back and forth as it relates to business intelligence. Right now it seems like operational efficiency is at the top of the list. And that's because we're, you know, we're going headlong into a sort of. Sort of a difficult economic period. And so organizations are looking to use business intelligence and analytics to, you know, to be more efficient. And I think that increasingly, you know, once again, having tracked this space for so long, organizations always invest in BI and analytics, whether it's good times or bad times. Sometimes there's a retooling that happens. So as we go into somewhat of an economic slump, we have to take stock and retool, focus on operational efficiency. But we're always going to be investing and business intelligence and data.

I like the idea that there's only two things in business you care about, so making more money or saving money. I'm curious as to whether it works. Was there anything in the report about whether these BI initiatives had had some success?

Absolutely. I think the majority of organizations would say that their business intelligence initiatives are successful. And the way they gauge that, I mean, some do look at things like ROI and. And I would encourage folks, you know, go and talk to your, you know, your finance folks and help them or have them help you determine the ROI associated with a particular application or initiative within your organization. Some departments and some folks do that, and I think that's really beneficial because that's a great way to get additional investment. But most organizations do get real value, and certainly sometimes it's just sort of this visceral sense of, hey, I'm better informed than I was before, or I'm able to take action with better perspective, much more questions quickly than I can before. So the greatest majority of organizations believe that their BI initiatives are certainly successful. Some would say very successful or extremely successful.

Okay, that's good news that it does actually work. Investing in BI is going to help you in the long term.

Well, and I would say if I. Let's take a step back for just one more moment and just take a look at the pandemic. And nobody wants to talk about the pandemic anymore. That was, you know, so 2000, right. 2020. But organizations that leveraged their data strategically and were well informed were much more likely to not only survive through the pandemic, which was, you know, fairly significant, negatively so for organizations, but some really thrived as a result of having that perspective. And I think that sort of the climate that we're in right now, where we have so many external forces operating our organizations, Data is your friend. You want to know when things are changing and how they're changing as quickly as possible so that you can react. I mean, at the end of the day, that's the whole point of business intelligence. How do we respond to change and execute as quickly as possible? And so I think that that's critical for organizations. And once again, going through the pandemic, those that did it, well, came out the other end and, you know, not that they were totally unscathed, but they, you know, they survived. Not everybody survived.

That's true.

Yeah.

The pandemic was certainly incredibly brutal on a lot of organizations. But I still, I like this ide bit of just business intelligence helping you be more agile in terms of your decision making. So I guess even more recently, we had lots of changes in US Policy around tariffs. Like there was a point in, in March and April where it was like every week there were about three changes. And so being able to react to those big things that affect you, I guess, yeah, that's definitely a dynamic.

And you know, once again, I mentioned the term hyper decisive. How are you able to respond to rampant changes? So all of a sudden, if there are tariffs, how do you reconfigure and find new suppliers, perhaps, or new channels or new customers? And I think that that's already happening out there. And once again, data is the key. If you don't have the data, you can't make those decisions.

Very true. So I want to go back to something you said a few months ago about how it's difficult to measure whether, like, how many good decisions you're making. So in theory, data is supposed to help you make decisions better. How do you measure this? I mean, I guess you get like, this is sort of proxy metric for, like, how many people got fired for making stupid decisions. But that feels like there's a lot of lag. It's not very useful. Is there a good way of measuring that you're making decisions consistently?

Well, well, a lot of it is, you know, based on sentiment. Right. It's subjective in nature. And if you, if you were, you know, to go to an executive, they would tell you that. Yeah, it certainly makes a difference. Of course, one way to know for sure is pull the plug and see if anyone notices. I think they probably would. Once again, organizations that are leveraging business intelligence, they do measure their success with it simply sometimes just by surveying their customers and their users. Sometimes they do use metrics like return on investment, where it's appropriate to do so. But I think they've become more scientific about that over time. And of course, adoption is another Measure, I'm not sure that's the best one, but certainly if you have broad adoption, that means the users are getting value if they're actually using it. So I think that those are typically the measures that organizations use to gauge success.

Okay, yeah, certainly. I mean, I guess adoption is a proxy measure, but at least it's relatively easy to track and reliably.

Well, I think, I think if there's resistance to it, if people don't see the value, and once again, you know, it's. These things tend to spread based on success and word of mouth. So if organizations or if individuals see this as key to their success, they'll adopt it. And so I think that deployment and adoption and utilization are reasonable metrics. They're proxies anyway.

Okay. Yeah, I think every organization's had this sort of problem where you've got a thousand dashboards and there's actually only six of them are used and the other ones aren't, aren't very useful. So I guess the things that are useful, yeah, they do stick around. So your wisdom of crowds report includes the idea of a strategic initiative. Can you talk about that? What's it involve?

Oh, so we, we, every year we add more topics to it. This year we have 65 different initiatives. So we basically survey our communities and ask them what's important to them, where are they investing, which things are critical, which things are not critical. And it's, it is sort of interesting to see, you know, at the top. What rises to the top are. Now historically it would have been reports and dashboards and visualization, but things like data quality, data integration have been really critical. And that's because once again, going back to the conversation of the fixtures versus the plumbing, organizations beginning to realize that we really need to focus on the plumbing to get this thing right. And as I said, there are so many other sources now, just not only internal sources, but external sources, cloud based sources that they have to integrate and to really get perspective, you need to leverage all of those. So we do ask about all these different things and we add new topics all the time. So I think this past year we added things like semantic layers, not a new concept by the way, but one whose time has probably come. And you see where it's sort of like the old top 40. You know, I, I don't know if you're, you know, a fan of that sort of stuff, but you know, where does, where does a new record enter in the top 40? And then how does it climb over time? And that's. Well, we have 65. So it's a lot more than that. But, you know, we see where they enter and then how they climb. So things like generative AI and semantic layer, they entered maybe in the bottom third, but I think they're going to climb quickly once they show their value. But once again, it's interesting to see that the mainstay of bi, the traditional stuff, is still at the top, including things like data warehousing. People like to talk about these new approaches, which are great. Talking about things like, you know, data lakes, and that's fine, but the old stuff hasn't gone away. It's still there, it's still being utilized. Organizations are still getting value from it. You know, I mentioned reporting. I mean, nobody wants to talk about reporting. I mean, it's, it's boring stuff, but it's useful. And so you, you know, sort of, as they say, horses for courses, right. You, you know, there are different approaches for different use cases for different Personas out there. And so you can't generalize. And so when we talk about things like conversational analytics and generative AI, sure, that's great for which use cases for which Personas, and that's where you really do need to have somebody who's in charge of this. And so for years I've talked about the competency center or a center of excellence or. And it doesn't have to be a formal organization, but there needs to be somebody who's thinking about this, somebody who is thinking about how do we approach this and documenting that and trying to achieve best practice. Nobody wants to talk about best practice anymore. But the reality is how do we do better? I think that things like whether it's loosely organized or more formalized competency center can certainly help to understand which technologies do we use and when.

Absolutely. So that's fascinating. I assume generative AI was going to be like top, because everyone's been talking about it nonstop for the last couple of years. So it is interesting that reporting data warehousing, these sort of real fundamentals of your data flow, are still the kind of the top initiatives.

There they are. And generative AI, I'm just looking at the list now, is pretty far down. Right. But it's, it's certainly important and I think that it will, it's going to blend into other things as well. You know, it's, that's the thing with any category. If it has any merit, it just becomes part of something else. And in fact, when we, we publish an annual generative AI report, it's not a Market. So it's just a generative AI report because it transcends multiple categories. And I think things do tend to bleed over into other categories and blend. And so I think that generative AI will be part of the equation and it'll be used in different ways to support other paradigms.

Absolutely. So it's maybe a complementary technology and I guess we talked before about it being sort of just thrust as features into other pieces of software.

But imagine if you will, and it's being done already. If I could simply use generative AI to, if I could reliably use generative AI to create a dashboard for me now that would be really useful because maybe I don't have the support for somebody to create it for me, maybe I'm not super skilled. That would be really helpful if, if I could in fact create a visualization and I could do it reliably and I could trust the underlying data and the analyses, I could see that as being valuable.

Absolutely.

Yeah.

Certainly just if you're designing, you go straight from I've made something in figma to this is a real life dashboard or even just I've drawn a sketch on paper thrown into some generative AI and suddenly I've got dashboard. That's kind of the dream there. One other thing you mentioned was the idea of a semantic layer. Now this is one of my favorite topics that I think people don't talk about enough. So talk me through. What's a semantic layer? Why do you want one? How is it complementary to AI?

Well, semantic. The concept of semantic layer is not new. It's been around for a really long time and I won't mention specific tools, but going back decades, in other words, putting a business view on top of the physical data and creating metrics and as opposed to, you know, dealing with the raw data. And most BI tools have some semblance of that in there already. But now there are many offerings out there that offer an independent semantic layer. So they can sit on top of any data and they serve as an intermediary between the data and the user tools, regardless of what the user tool is. So we're dealing all dealing with the same metrics with the same business semantics. And the semantic layer figures out how to generate the data manipulation language of SQL behind the scenes. And they do caching and things as well. And I see that as being highly beneficial. We, we do have a plethora of tools out there. The larger the organization, the more tools they have out there. And they have their own semantic layers and they generate SQL in their own special ways. I think it makes sense to have all those talk to a semantic layer and have the semantic layer be that intermediary so we all have a consistent view. And so like I said, it's been, the concept has been around for a long time, but I think it's one whose time has come, especially as we look at AI based applications and we want to make sure that we have the appropriate context for those applications. So I'm excited about the concept of an independent enterprise wide semantic layer.

Absolutely, yeah. Having different teams calculate like the annual recurring revenue or whatever in different ways. That's going to be a disaster. Any kind of decision making.

Well, we've always seen that. Right. You go into a meeting and finance have their numbers and sales have their numbers and they think they're, you know, talking about the same thing and they're, they're not. And not to say that it's easy to create those semantics and those KPIs, but it's really important. It's really a business activity. It's not a, I mean obviously it's technology associated with it, but it forces the business. Just like when businesses implemented erp. Remember that it was really hard because they had to go through that process and we have to do it again. If you spend the time and craft a really meaningful and useful semantic layer, once again it helps to achieve that notion of information democracy and hyper decisiveness. So plumbing, I know, not sexy, but really essential.

Okay, this seems to be a recurring theme is like you've really got to get the basics right in order to have any success.

All right, how many of you want to drive a few fancy cool looking car that is mechanically and electrically bankrupt? Oh that's right, that. We already did that. It was Leland, right, British Leland did that for years. They don't exist anymore. So.

No, it's been a good fortune.

Yeah, we owned the Triumph years ago and really sexy cars when they ran. Anyway, we don't want those. Right. We want the car that looks great but also is mechanically and electrically excellent.

Absolutely, yes. I like the idea of mechanical excellence in your, in your data stack. So I'd like to talk a bit about budgets. I mean, so we talked a bit before about how, yeah, there's sort of economic uncertainty in many parts of the world. How has that affected BI budgets? So are they going up? Are they going down? Are people investing more?

Well, I would say that this sort of leveled off a little bit. Most organizations are investing or they're holding. Very few are. I would say Reducing their investment. But I would say it's not as bullish as it was even a year ago because of that uncertainty. But there's still investment. In some cases what they're doing is they're pulling budget from other places and they're investing in business intelligence and the supporting data. So still positive. But like I said before, there's always a retooling that's going to take place when we end up hitting some of these external forces that are negative. And that's definitely where we are. But right now, I would say during difficult times, business intelligence becomes more critical than during good times. And I think that's what we're seeing right now. It's that retooling. So investment is still really strong, but perhaps not as strong as it was a year or two years ago.

Where are the costs coming from in bi? Is it just software licenses or is there more to it?

So most of the cost is people. I think things have certainly shifted over the years. It used to be perpetual licenses and well, that's gone. It's all subscription. It's mostly cloud based and there's still a lot of on prem out there. That's kind of the dirty little secret that the software vendors don't want to talk about because they're paying maintenance right on those things. But there's obviously strong initiatives from most of the vendors to move those customers to the cloud and to a subscription based license. And so there's been that shift to obviously subscription based licensing. But people still are at the top of the list. That is the biggest line item. And hey, you still need people. Even in the world of AI, you know, you hear stories about organizations laying off entire departments. Don't do that. That's a bad idea. They're not coming back once you let them go. And especially with business intelligence and things like data engineering, we talk about things like semantic layer. You need really capable, really smart people to do these things. And so we still need people. That's still going to be your largest line item for the foreseeable future. Maybe it even has to grow. I'm not sure.

Okay, so that's interesting that people are kind of. I suppose it makes sense. People are more expensive than software licenses. That's the biggest area you need to invest in. But I'm curious on this idea that there's still a lot of organizations running individual licenses on their own computers rather than a cloud software. Is it like, do you have a sense of like what the fraction of people just still rocking a copy of like Excel 2013.

It's, I would say it's, it's a, it's a massive population of folks and that presents obviously a problem for, you know, software companies. But I'm not shedding any tears over that. If the users are getting value and that's why we, when we rate vendors and look at markets, we don't look at just only cloud based. Well, no, there are a lot of folks that are still using on PREM software. So much work gets done on PREM these days and cloud certainly has real advantages and that's where a lot of the investment is. But there are tremendous populations of users that are using desktop tools out there and still getting value. And that's, I don't think that's going to change. And you know, I've gone to a number of vendor conferences over the last couple of quarters and many of them saying, don't worry, that's, it's not going away, we'll continue to invest because they know that those customers are not going to move to the new architecture or to the cloud for that matter, and they need to continue to support them and they're obviously their valued customers. So I think over time they'll move to the cloud. But as long as they continue to get value from their desktop or on PREM software, it's going to take a while.

Okay. Yeah, I guess I'm always surprised at like the length of time it takes, but as the shell companies like using mainframe computers.

Never went away.

Yeah. So you mentioned the idea of rating vendors. You said it's more than just about.

Okay.

It has these features at this price. Talk me through. What's your process for rating vendors? What should people care about?

Yeah, so we, it's all driven by voice of the customer. So we actually collect data around the. We have these 33 measures that we've been using since the beginning which evaluate the vendor and the product, all of the customer facing activities. So sales, technical support, consulting services. We added TCO as a new measure in the last couple years as well. And so we collect all the data from the users, from the buyers, the people that are in the best position to evaluate a vendor and a product's performance. And, and that's what we use. It's, it's very simple model. It's not magic at all. It's very transparent in nature. Okay.

So I like the idea of just thinking about the sort of broader software experience. It's like, is it easy to buy? Do you get good support, all that kind of stuff.

Yes. And then the product Looms large there, too, you know, in terms of. And these, you know, the criteria that we use are at a relatively high level. But does the product scale, is it, you know, is it integrated well with other technologies? All those things? I think. I think we have 12 different product measures that we use, as well as all the other customer touch points. We also ask about the perceived value for price paid, the integrity of the vendor, whether or not they'd recommend the vendor to somebody else. So all those things come into play when we evaluate the vendors. And then we use three collective models, and then we have dedicated models for each vendor. Because I believe, you know, you know, I was at Gartner for 13 years, so I know all about MQs and how they're managed. And I think a lot of assumptions have to be built into that. And we have multiple models because I think it makes it easier to understand the marketplace and the vendors. If you, if you have that.

Absolutely. So can you tell me, is. Is the BI space, is it getting better on, like, which of these 12 metrics are things improving on? Why is it getting worse?

I think products, for the most part, I mean, products have come a long way. I think products are very usable. They're easier to deploy and all of that. I think where the vendors lose marks is because of all of the M and A activity, private equity investment, trying to, you know, cut costs, lots of reorganization. We see that in the numbers mostly on the sort of the people side of things, because they. There's, you know, there's discontinuous change. People don't like that. So all of a sudden, the people who are dealing with the four are gone. So whether it's a sales rep or a consultant or whether it's tech support, I see that as problematic. So it's that discontinuous change that customers are having an issue with. But the products have come a long way. Of course, there's been a disproportionate amount of focus on things like integrating generative AI, and some of it will be used by customers, some of it will not. So maybe a little bit of a distraction in that regard. But I think that the state of technology is pretty good and pretty robust and has been making great progress over the course of the last 10 years.

What I'm getting from this is people are still complaining about customer support. I feel like this is true for, like, any bit of software is. Yeah, stuff's changing.

When we do our regular surveys, we see it. It shows up in the numbers you could. Same way you can see when a new product is released and there are product problems where it shows up in those 33 measures. It's a really good barometer to understand what's going on with, with the vendor. But, yeah, I think that's been largely the case and hopefully that's going to settle down over the next 12 months. We hope so.

So you mentioned that a lot of the products have improved over the last 10 years. Are we at the point where BI just works or is there still more to do?

Well, there's always more to do. Right. But I think that the greatest impediment has been probably cost associated with it. We see lots of vendors raising their prices and that limits, I think, deployment. But I think from a, from a usage perspective, yeah, these things are very usable these days. And once again, you know, AI can help with some of that too, increasing the usability and maybe increasing time to deploy capabilities. So I think that, you know, I don't want to say the job is ever done. I think it's more of an evolutionary path. We are so much better off than where we were 10 or 20 years ago. But there's always room for improvement.

Okay, yeah, definitely room for improvement. But it does sound like we're at the point where, you know, things kind of mostly work. Yeah.

Oh, yeah. There are some great products out there. There's no question. I think that. And I don't want to say all products being equal, but you go to the, you know, to a conference sometime, you go to the show floor, they, you know, they all basically do the same thing, more or less. So, you know that if a, you know, if a product has been out there for a while, supported by the vendor, you know, it's going to have certain capabilities and you're going to be able to get from point A to point B with that product for the most part. There are some variations architecturally. Sometimes there are things that are worth taking a look at to see how they've architected the product. The ecosystem is increasingly important. So, you know, I was just at Databricks last week, and so there are a lot of different ecosystems out there. There's the Snowflake ecosystem and the Databricks ecosystem and, well, you know, the rest of them out there too. The Google ecosystem and the Amazon ecosystem. And so sometimes that matters to an organization, what ecosystem they're a part of and whether or not they're in that vendor's marketplace and many other facets that one, you know, might want to consider when selecting vendors.

It seems like there's lots of great choices out there. I'm curious as to how close we're getting to self service analytics. This seems like it's the long been the dream that the products just work even if you don't have a background in data, you just want to create your reports, create your dashboards and be done well.

And so self service has been limited because it required a lot of IT support in order to do it. And that's gotten better certainly. But you know, it have other priorities out there and so if you want changes. So all right, so now I have a dashboard and that's great and I can even, you know, modify things and create things in my own little environment. But if I need anything added to it, I have to go back and get in line. And so I see over time, once again, you know, maybe this is where things like generative AI can help, where users can do more for themselves. So maybe, maybe we're on the cusp of seeing much greater success. I think those that have self service capabilities are being relatively well served. They're getting value from it. But once again we talked about the information underclass. How do we get it to them? And maybe it's not through the traditional approach of self service. Maybe it's a new approach.

Okay, yeah, I guess the traditional approach is like it's self service after a data engineer spent a few hours writing a SQL query for you. And yeah, now it's like maybe generative AI, you can write it for yourself.

I mean it's better than the days of, I don't know if you remember the old days of executive information systems where they were put together with bail, wire and glue and required an entire staff to support the executives. We've come a long way from that. I mean it's so much better and so much more deployable than it was all those years ago. But once again, like I said, there's always room for improvement. So how do we get beyond that 40% where we are today and we can call it, I don't care what you call it, you call it self service, call it something else, doesn't matter. How do we get to even 50%? How do we get to 60%? Because the, you know, the value to doing that would be tremendous.

That sounds like an incredibly important question. Do you have an answer? How do we get there?

Well, I do think that things like generative AI can once again, if we take into consideration all the governance aspects, can certainly help because it can get to individuals that may not be as data savvy. But still need access and don't want to go through a dashboard. They just want to ask a question because maybe they're not using it every single day. They're not like, you know, with a dashboard. People that, you know, rely on their dashboards, look at them all the time, tracking key metrics and have important questions to ask. Sometimes I might just want to ask a question at the end of the quarter and that question may not be a question that it has anticipated. So I do think that there is value for things like conversational analytics and generative AI once again with proper governance and security in place to more fully deploy business intelligence capabilities in the enterprise. And maybe that's what increases dependent. I think it has the promise of increasing that penetration. You know, will we ever get to 100%? Not likely, but we should strive for it. I think it's important to strive for that.

Absolutely. And it's really interesting the idea that, like, maybe you don't have a question for months, then it comes in the quarter. I have a question now. I need the answer immediately. So then you're like into, well, it has to be done. Sharpie. She got the really fast turnaround time.

Absolutely. You need it. I need it right now. Exactly.

All right, super. So just to wrap up, what are you most excited about in the world of business intelligence?

Gosh, once again, been in the space for a really long time. It's the gift that keeps giving. It's a really exciting time. And, you know, AI certainly is part of that. Things like semantic layers. There's so much innovation going on and also has become front and center. I remember a day when applications were built without considering how you would analyze the data. There was no provision for analyzing the data, which now is crazy. Nobody would ever do that. And so data all of a sudden is king. We always are thinking about data and we're seeing a shift away from the application centric model to a data centric model. Organizations understand that now. So it's a really exciting time. Very exciting to see us really focus on the value of the data. So that's sort of unprecedented. We haven't seen that until, you know, the last few years.

That's actually very cool, the idea that you're not just collecting data for the sake of it, you are actually thinking about, well, what data is going to be useful? I can analyze this and then get some sort of impact from it.

Yeah, it's become truly strategic in nature. And that's why you have chief data officers and chief analytics officers, because there's that, you know, heightened awareness that this is the gold within the organization. We have to mine it.

Absolutely. Wonderful. All right. And just finally, I always want recommendations for people to follow people's work to look out for. Whose work are you most excited about at the moment.

From a research perspective? And I have to say our own. No, I mean, I, I, you know, obviously we're friends with all the other analysts, and they all do great work out there, but, you know, truthfully, we've been, you know, we've had a focus on primary research for the last, you know, going on 18 years now. So our, our 18th birthday is coming up next month. I think that tracking the progress of the industry and understanding where the industry has been is critical to understand where it's going in the future. And so I think anyone that's focused on primary research, I think that's really worth looking at because it avoids all of the hype. So not just ours, but, you know, any primary research I think is worth reading about to understand that historical perspective and to sort of, you know, gauge your own initiatives based upon that, as opposed to those that basically want to, you know, focus on the new shiny object. I'm not a fan of shiny objects. I'm a fan of being more pragmatic and understanding where the value truly is for organizations.

Yeah, certainly there's definitely plenty of shiny toys to get distracted by at the moment, and some of them are fun to play with, but, yeah, you've always got to bring it back to, are you getting some value out there? All right, super. So thank you so much for your time, Howard.

Thank you, Richie. This has been fun. Sa.