When it comes to business intelligence, Howard Dresner is the expert to speak with. Dresner is credited with coining the term ‘business intelligence’, and has been a thought leader in the BI field ever since.
His firm, Dresner Advisory Services, has recently published its 7th annual report on the topic, “2018 Cloud Computing and Business Intelligence Market Study." At 102 pages, the report is the deepest of deep dives into the rapidly growing Cloud BI sector.
In a wide ranging conversation, I spoke with Dresner about what the report indicates for the Cloud BI sector, now and in the future. The following interview has been edited and condensed.
Howard Dresner, Chief Research Officer, Dresner Advisory Services
1) 2018 Is a Breakout Year for Cloud Business Intelligence
In the chart about importance of Cloud BI between 2012 and 2018, there was a sharp spike upward. In particular, there was jump from something like 30% to almost 50% between 2016 and 2018. It seemed like 2018 is kind of a breakout year for Cloud BI. Why do you think this is so?
 was sort of a watershed event. I think that it’s gotten to the point where organizations have gotten through the knot hole. I think some of their tentativeness or fear or concerns about cloud have been allayed, in large part. Even finance seems to be more comfortable.
For a long time, FI and IT were pretty closely aligned in their resistance to cloud based deployment. Now we’ve seen those two sort of diverging. IT, in general, is still more reluctant toward cloud. I think that a lot of it is loss of control. It is a fundamental change in the nature of their job and their organization. So, it’s disconcerting and it’s a much more disruptive change than for finance, let’s say. Moving from the server under the controller’s desk to something in the cloud. It really doesn’t change their job that much. But for IT it does quite a bit.
So, I think that’s in large part what’s happened. I think that especially within smaller organizations there is virtually no resistance to moving to cloud. There are so many offerings now that are in the cloud; many of which are cloud-only or cloud-first.
It makes much more sense for organizations to go right into a cloud-based deployment or to migrate to a cloud-based deployment, because that’s where you’ll likely have more choice, and also more likely to have newer features. Even though those that have on-prem software will continue to maintain that software, it’s not going to be on the same schedule as the cloud-based offerings.
2) Feature Requirements for Cloud BI are Similar to On-Prem Requirements
There’s chart that talks about Cloud BI feature requirements. I’ll quote briefly from the report: “In 2018, conventional BI functionality, including dashboard advanced visualization, ad hoc query, data integration, and self service lead the list of the most required Cloud BI features.” It sounds like there's a lot of overlap, if not a total overlap, between what people want on-prem and what people want in the cloud. Is there a big difference there?
Not a substantial difference. What’s interesting is that there has been a shift. So, for a period of time on the supplier side of things you still had to use back client, or design for administration, or things of that sort. A superuser. And that has changed, as well.
So, now you have much more of a 100% pure cloud-based implementation. But from a functionality perspective, yeah, in large part they want the same thing, with some differences. If you look at on-prem, the number one on-prem paradigm would be reporting.
Now, when you’re looking at cloud based implementation, you’re looking at something that is much more dynamic and interactive. It’s a little different, but there’s significant overlap.
3) Cloud BI and the ‘User Revolution’
The really interesting point here: 2018 saw a surge in interest in end user self service, which to me says there’s sort of a democratization of the technology. It’s not just the folks in the c-suite, but the folks in the cubicles and the sales reps. You don’t need IT to do this, you can do it by yourself. It seems like that’s a big trend. Would you agree with that or disagree?
It’s all depends on how you define self service. To the average end user, self service is ‘let me take my own tool and get out of my way.’ If you talk to IT, it’s ‘we’re going to choose the self service tool for you in a governed environment and we’re going to determine what the governance is.’
There’s probably a happy medium. I think that yes, users want to have freedom to analyze the data, to combine it with data that they find is interesting or valuable to them, and not have to go ask somebody. Because, when you have to go ask somebody that introduces delay. That’s the problem. Very few IT organizations can be that responsive. So, I think that the trick is getting the users those things that the users ought to have, while retaining some control over those corporate assets that do need to be highly governed.
That said, do I think that there’s been some sort of a user revolution? Absolutely. There has been for a while. Many times, the users are acquiring these solutions, sometimes with the recommendation, with some input from IT, but IT is not the economic buyer anymore. So, IT definitely has some sway in terms of influence, and in terms of applying or imposing some modicum of standards. But they can no longer control it, and they may not even have veto power.
It’s the democratization effect. Go back a million years ago, in 1993, they started taking about this notion of information democracy. Part of the problem was that cost prevented information democracy, and as the cost comes down and as these tools become generally available and more accessible, where the business user can do a lot by themselves, you get a lot closer to that vision of information democracy.
4) Cloud BI and Security
There’s an interesting chart about Cloud BI security requirements 2017-2018. The report says respondents, ‘found an even greater gap in awareness of security standards than in our past studies.’ I’m think that one of the stumbling blocks to Cloud BI is, ‘oh, we’re not sure if it’s secure. We don’t want our data out there. We want to hold onto our data.’ So, am I reading this to say, you know, there’s even less awareness of security than there was in the past? What’s happening to security issue in terms of BI?
It’s kind of a paradox there. There paradox is… and you also have to pull it apart based on who’s answering the question. So, if you’re in sales you probably don’t have a lot of awareness of security. You have AWS cloud. You’re not going to worry about it because you assume that they know what they’re doing. Or you’re going to be in Azure or Google’s cloud, so you’re assuming, and this is sort of my take away: I don’t need to know because Amazon knows. You know what, we don’t have to focus on these security protocols ourselves because our supplier does that for us.
That said, when I hear IT saying that there are issues with security as a rationale for not investing, and then they say, well, we don’t know a lot about security, that’s a problem. But, I think it’s more the former than the latter. People just assume that one of the big providers, who are going to be hosting the application, are taking care of that for them.
[The big cloud vendors] have to spend millions on that and have many highly skilled people that are constantly monitoring and innovating to keep the bad guys out and to prevent that from happening. All it will really take is one, as we’ve seen in other organizations: one big breech to lose the trust. If you’re Amazon or Microsoft, you can’t afford to have that happen because you’re going to have significant impact. So, they spend a huge amount on people and technology to try to stay ahead of the bad guys.
5) Will Cloud be Dominant in Business Intelligence?
A couple of global questions to wrap things up. Reading the report and just seeing the market as I see it, long term, not just the next couple years but long term, I think 5 years plus, cloud is going to be the dominant model for BI. The idea of on-prem BI is essentially going to be obsolete. Agree or disagree with that?
I agree with part of it. So, will it be dominant? Absolutely. It’s becoming that way. We’ve been publishing this particular report for seven years now. That’s why I have certain favorite charts in here, because we have seven years worth of data. It’s really cool. You can see real shifts. If you look at Figure 19: Plans for public Cloud BI 2012-2018. Wow. I’m sure those people that said ‘no plans’, the 70%+ in 2012 never thought that they’d be deploying by 2018. Many of them did. So, I love looking at charts like that. Being data people, we love data. More data is better. Will we do an eighth year? Probably. But, at some point we’ll discontinue it simply because it’s going to be the de facto model.
That said, you know there are still 3070 terminals out there, James. There are old style TRTs out there running CICS against IMS on mainframes. Am I speaking in a foreign language? I’m an old guy.
Everything persists forever. It’s just a matter of quantity. There’s still vaxes out there running stuff and they haven’t been replaced because it works. They’re not going to get replaced because of that. There’s nothing that will do it as effectively or cheaply, so they just leave it alone.
I do think that as new applications and new users come online, it will increasingly be public cloud-based. I’m comfortable in saying that. It doesn’t mean that everything else is going to get displaced. Some will.
What’s interesting to note: we asked a new question this year and we’re going to publish the results of this. This, we haven’t published yet, but we asked what people rated a vendor that they were using. Did that product replace another product?
Only about 23%-24% said that it replaced something existing. If you look at the plans for growth and the budget growth and the fact that in a lot of circumstances they’re not replacing something, it says that there’s a lot of opportunity to automate users that don’t currently have access to anything except maybe Excel.
6) Cloud vs. On-Prem Business Intellligence, Long Term
It seems like the eventual problem with on-prem is going to be that it won’t be able to keep up with the tool set because the cloud-based BI can upgrade a tool set overnight whereas that on-prem dinosaur, it can’t do X, Y and Z anymore and people will go, wait a minute, it can’t do X, Y and Z? Do you think that’s going to happen or not?
For some it will. There are people still out there, once again, I’m showing my age, there are still people out there using nano core systems from the 80s. You know what? You’re not getting new features.
People get comfortable with something and even though they might have gripes about lack of functionality, it’s the devil they know versus the devil they don’t know. So, I do think that gradually it’s going to be a much more gradual process than people would suspect. So, new applications, new users, makes perfect sense to go with the new thing. And for those who are already existing users, those that want to move will move. Those that don’t probably won’t.
So, I think we’re going to have coexistence for a lot longer than probably people think we will until somebody wants to wholesale migrate to the cloud. In some classes that happens, too, when people start talking about re-platforming. Okay, so that might be part of the re-platforming discussion. But, you have to, to some degree, you have to moderate the industry hype out there. I mean, it’s not binary. It’s not, okay, you know, now it’s on and moments later it's off.
Markets don’t work that way and users don’t work that way. They’ve got some costs and things, and it’s going to be a period of time before we see that kind of majority of existing users migrate to the cloud. It may happen eventually, but it’s certainly not going to happen next year.
7) Key Takeaway: Cloud Is Preferred Model for BI
Do you have kind of a global takeaway from the report?
Well, I would just stay that, increasingly, cloud is the preferred way to go for many people. It’s less costly. There are many more options out there. Many organizations and suppliers are taking a cloud first approach or cloud only approach. Especially if you’re dealing with business partner or user buyer, it’s sort of a fait aucompli. That doesn’t mean that desktop software is dead, by any means, but there are so many attractive options in the cloud that it certainly mandates careful attention for someone who is selecting a product.