Dresner: Top analytics trend remains rate of adoption
Increased investment in analytics and proliferation of data science capabilities are among the most significant ongoing trends in BI as the pandemic stretches into late 2021.
With the pandemic well into its second year, a chief ongoing analytics trend remains the recognition of the importance of business intelligence as a means of managing uncertain economic times.
When COVID-19 first began to spread in early 2020 and governments responded by issuing stay-at-home mandates and essentially shutting down local economies, many organizations turned to analytics as a way to figure out how to survive.
They were in crisis, and data helped them understand what was happening.
Now, while the pandemic remains a health crisis, among the enterprises that were able to survive, many have fully reopened and recovered.
And according to Howard Dresner, founder and chief research officer of Dresner Advisory Services, organizations have not stopped using analytics to their advantage. But instead of relying on data reactively in the face of uncertainty, they're using it proactively to guide growth initiatives.
The growing reliance on data, however, is not the only significant ongoing analytics trend, according to Dresner.
In an interview, Dresner discussed the current analytics landscape, including which trends he thinks are underrated and which are overblown. In addition, he spoke about what analytics trend has most surprised him this year, and what still excites him about data three decades after he first began as an analyst at Gartner and nearly 15 years after he founded his eponymous firm.
As we enter the final few months of 2021, is there any one analytics trend that has stood out from the rest so far this year and been most significant?
Howard Dresner: One is accelerated investment in data and analytics, and that's a result of the pandemic. Even if organizations weren't increasing overall budgets, they were reallocating budgets in favor of data analytics, performance management and other business intelligence stuff simply because during crisis you have to get your head around what's going on, and you have to do it as quickly as possible so you can make decisions and execute with precision. If you don't know the data, then you don't know what's going on in the outside world or what's going on in your business and you make bad decisions.
Things like across-the-board cuts are always a bad decision. When I see companies do that, it tells me they don't have their act together and they don't know what's driving the business. They don't know where they ought to be investing and where they ought to be cutting back. We've been collecting data around COVID since March of 2020, and things are actually improving a little bit in the sense that things are stable, people know what they need to do in this pandemic environment, and they're starting up more analytics projects. We're seeing more organization start to invest, and they're hiring people. If you're a data person, you're in high demand. That's a positive thing.
What is another noteworthy analytics trend this year?
Dresner: Everybody wants to talk about data science. Data science and advanced algorithms are not new, but what's different is our capacity to actually leverage these things. For core data science technology and processes, it's still largely an endeavor for big organizations. Large organizations have data science departments. Within large organizations, we're seeing more process around data science. ModelOps is a real thing. Not everyone is doing it, and not everyone who's doing it is doing it well, but it's something you typically see in large organizations.
For smaller organizations, data science and machine learning algorithms are showing up in their applications. You won't see a data science department in a small or midsize enterprise -- they just don't have the resources to do that -- but whether it's their ERP applications, performance management applications or their CRM applications, they all have smart features built in, whether you like it or not.
With respect to those smart features, do users actually like them and want them?
Dresner: Depending on the function and the industry, there are varying degrees of acceptance. If you talk to senior management -- the CEO and the CMO -- they're much more receptive, but if you talk to the CFO, they're not as receptive. I don't want to say [the smart features] are 'Big Data 2.0,' but it does kind of feel that way with so much hype. It is much more vendor-driven than user-driven. Depending on how you slice the user demographics, there are some hot spots and really good use cases, but the enthusiasm on the part of the vendors is much greater than the users.
Howard DresnerFounder and chief research officer, Dresner Advisory Services
Some of the recent announcements around vendor valuations speak to this. They've created this dynamic right now in the marketplace where there are these tremendous expectations around the impact this technology will have on the future. Maybe it will, or maybe it's Big Data 2.0. I think the advanced technologies -- AI, data science, machine learning -- definitely have a home, and there are real use cases and there will be more use cases over time, but I don't think they're going to take over the market in the next five years.
Last year as economies shut down in response to the pandemic and businesses were sent into crisis mode, it accelerated trends like digital transformation and analytics adoption. Has that momentum continued in 2021?
Dresner: I think the momentum has continued. I see that the mandate to do more with data is maybe even more critical this year than it was last year. In March of last year, everybody got a huge wake-up call. It was one of those moments when the sky was falling. A lot of companies got shut down. There are still some companies that are shut down -- not too many -- and there are some with divisions still shut down. What companies are telling us when we ask about loss of business, loss of revenue and employee productivity is that the number talking about loss of business and loss of revenue is down, so they're recovering. And we're seeing lots of new analytics projects, more than twice as many as when we measured in the spring of this year.
Now that things are a bit more stable, folks can think more strategically. Last year, it was completely tactical. It was reactive. Now, folks who are still standing have a more stable environment, the worst is behind them and they can make strategic investments. They also know there will be another crisis, so if there's anything healthy that's come out of the pandemic, it's that it's forced organizations to rethink how they do business, and how they understand the business.
Are there any analytics trends that have surprised you in 2021?
Dresner: I expected cloud adoption numbers to get blown out of the water, and it didn't happen. Cloud adoption is definitely happening, but it's been linear in its growth and adoption. We expected to see a huge bump, and it didn't happen. It's been in step with its previous trajectory, and that's because to move to the cloud you have to spend money and resources. That kind of a sea change is more strategic than tactical, and given the climate we've been in, if organizations already had plans to move to the cloud they were going to continue to do with those plans, but they weren't going to create new projects in that climate.
That may change next year. But that was surprising. I thought everyone was going to go to the cloud, and that hasn't happened.
Are there any analytics trends you think are overhyped?
Dresner: The advanced stuff -- data science and machine learning -- that's overhyped. There's real value there. We use a lot of tools that have AI baked in and they're really valuable to us, especially things like natural language processing. For example, when we do vendor briefings, everything gets recorded and transcribed using AI. We're using some other things out there for more advanced things, like GPT-3. But it's early days [for those capabilities].
They have promise, but I think they get overhyped. There are a lot of vendors trying to get above the noise, and the way they get above the noise is to say, 'It's this shiny object you need to pay attention to,' but the reality is you still need to pay attention to the basic blocking and tackling of data and analytics. You still have to do the work, and you have to have people who know how to do the work.
What about an analytics trend that's perhaps underrated?
Dresner: One of the things we talk about a lot is data literacy. Data literacy is critical. Just because we have data science and advanced algorithms doesn't mean that the rest of the organization can be ignorant. It can't. And I would argue that it's more important that people be fluent in data now than ever before. Some organizations are clearly investing in that, developing and implementing programs to raise the data IQ of the organizations.
Have there been any analytics capabilities that one vendor or a group of vendors have developed this year that has stood out to you?
Dresner: No, not really. There is some interesting technology out there, for sure, but I think we might be moving into another period of time when people are looking for comprehensive, organic functionality. In other words, the stuff all needs to work together as opposed to a multiplicity of tools. It's split in our research community. About half want an organic, integrated platform soup to nuts -- integration, data preparation, analytics, user interface, embedded BI. And then others want to build their own stack. Every few years we go through a cycle where best-of-breed is in and then we shift back to remembering why we didn't want to do best-of-breed because it's really hard to integrate everything.
I think that bodes well for vendors that haven't done a lot of acquisitions. It's easy for vendors to buy stuff to get new functionality and get new talent, but then the integration effort that ensues is really hard and takes years. You can do very basic integration, but deep integration with common architecture, common data structure, common UI, common admin is hard.
So do you still favor a lot of the traditional, proven capabilities?
Dresner: I'm not a fan of shiny objects. The top paradigms remain the ones that were in place last year, the year before that, and the year before that. When people say dashboards are dead -- hogwash. Dashboards are not dead. They're as important as they ever were, and it's the same thing with reports. None of the vendors like to talk about reporting, but that's the top paradigm -- just send me a report. It can be a PDF with links in it, but there are a significant number of users out there that don't want an interactive experience. They just want to be sent a report, look at the report and then move on with the rest of their day.
The meat and potatoes of BI and analytics remain relatively constant, even though there are some things that are climbing up, like the cognitive capabilities and things of that sort. But they're still at the bottom of the list. [Visualization] is still in.
What's going on in analytics that excites you right now?
Dresner: Everything. We're looking at a lot of things and constantly expanding our agenda. This year, we're collecting data on things like people analytics. We started covering natural language analytics. On the finance side, we're covering financial consolidation, which we've never covered before. We're starting to see functional application areas as significant and important, ways in which technology is entering the doors of the organization. But I get excited about anything that has to do with data, whether it's data we're collecting or whether it's products that leverage that data and deliver value to the end consumers.
Finally, are there any analytics trends we didn't talk about yet that you'd like to mention?
Dresner: The digital enterprise is happening, and those that don't figure that out are going to be at a significant disadvantage. And one of the really cool things about the digital enterprise is that it generates a lot of data. Everything generates data, and everything can be monitored. You can have sensors everywhere -- that's the good news and the bad news because they generate lots of data and not all of that data is signal; a lot of it is noise. So much of this has to do with people being able to quickly, by having processes in place, leverage the insights that are derived by the data and then act upon them. We see that as pretty significant moving forward into the future.
Editor's note: This Q&A has been edited for clarity and conciseness.