5 Tips for Getting Started with Advanced Analytics

Loraine Lawson

Updated · Oct 06, 2016

Advanced and predictive analytics seems to be at the top of its game, at least when it comes to hype. Ninety percent of companies rated advanced and predictive analytics at some level of importance in the 2016 Advanced and Predictive Analytics Market Study, which is conducted annually by Dresner Advisory Services. It lands at number six on the firm’s list of 30 top tech trends, ahead of cloud, governance and even Big Data.

Here’s the rub: The survey found the adoption rate for advanced and predictive analytics was 24 percent. Howard Dresner, chief research officer for Dresner Advisory Services, believes there’s little that indicates the number will grow much will in the near future.

“It’s in the top third of priorities for all of these related technologies and initiatives out of 30 that we track, so I find it interesting that there’s a degree of urgency but it doesn’t map to the adoption,” he said. “So people think it’s really important but, yeah, we’re not going to do anything about it. That’s the truth.”

[Related: Check out our buying guide for predictive analytics software]

Nearly 50 percent of respondents rated advanced and predictive analytics as “critical” or “very important.” Only 10 percent rated it as not important, with the rest classifying it as important or somewhat important, according to a press edition of the report.

Another surprising finding: Large and small organizations are the most likely adopters, leaving mid-size organizations behind. Small companies are more agile and it’s generally less expensive for them to act on analytics than larger companies, Dresner said. While large enterprises may take a bit longer, they have the resources to invest in advanced analytics.

“The guys in the middle just get a little bit stuck,” Dresner said. “It’s harder for them because they’re certainly not as agile as the really small companies and they don’t have the resources of the big companies. You see this in technology in general.”

The Dresner report isn’t the only one to raise questions about the hype versus the reality of advanced and predictive analytics. Bain & Company’s Advanced Analytics Practice group notes only four percent of companies are able to combine the right people, tools, data and organizational focus to take advantage of Big Data and advanced analytics.

[Related: Read about our data preparation tips for predictive analytics success]

What’s the forward-thinking technology leader to make of such findings? Here are some tips to help reframe your thinking on advanced and predictive analytics:

Don’t Dismiss Advanced Analytics as Pure Hype

Despite these market studies, it’s hard to actually know where companies are in terms of adoption, said Sarah O’Brien, practice area director for Bain’s Global Advanced Analytics Practice. That’s because adoption can be uneven within industries and even within individual organizations. Insurance is known for its use of advanced analytics in its actuarial divisions, but the use doesn’t always extend beyond that division, for example.

“It is difficult for anyone to pinpoint specific adoption levels and trends because advanced analytics covers a wide spectrum, from deep predictive and prescriptive modeling to more traditional methods on larger and more various data sets and incorporating more unstructured data,” O’Brien said. “I would say while the bleeding edge is something that is challenging for many companies to obtain given their existing capabilities, there is a spectrum — and we do see companies across all industries looking at how they can improve use of analytics.”

That said, companies that jumped at the hype without an advanced analytics plan have failed, warned both O’Brien and Dresner.

“You see organizations that have failed with this stuff; they see it as a panacea. It’s like, ‘OK we’re going to do advanced and predictive and find the needle in the hay stack,'” Dresner said. “That doesn’t happen too often.”

Develop an Advanced Analytics Roadmap

“It’s going to be more of an evolutionary process,” Dresner explained. To help guide that evolution, organizations should develop a roadmap for advanced analytics.

For many organizations, that will require an overhaul of the data infrastructure. Bain’s IT Practice found that 85 percent of companies say they need a substantial upgrade to their data platforms. A roadmap to analytics may mean replacing legacy systems with cloud-based solutions that support scalability, or replacing ETL with data management processes that support a move to real-time analytics, the strategic management consultancy advised.

Even so, successful companies focus on baby steps rather than wholesale rip-and-replace projects.

“We see companies instead of trying to tackle the all-knowing master data management, multi-year process, we see companies picking off very specific business use cases — starting frequently in their digital channels, where it’s often easier to corral various sorts of data and investing in just the minimum amount of data management infrastructure in order to facilitate proof of concept in a much more narrow band and then moving out from there,” she said.

Another key step on the advanced analytics roadmap: finding success with business intelligence. Organizations that are successful with BI are much more likely to adopt advanced and predictive analytics, found the Dresner study. This is likely due to the fact they’ve already cultivated mature data practices, Dresner said.

“Organizations need to have a plan or a roadmap for this because there is real value to doing this and it does inform a lot of the other activities in the organization,” he said. “But it’s not going to change your lives overnight, so I think that organizations are more focused than ever before on some of the basics; I think that’s what we’re seeing.”

Don’t Expect Vendors to Solve Advanced Analytics Issues

While vendors are certainly striving to democratize data and advanced analytics, both experts warn organizations not to expect vendors to solve their problems.

“I think it’s not a sufficient answer for all companies and all decision areas because I think it then takes a lot of the ability to learn from the analytics and deepen the company’s own capabilities away from the buyers,” O’Brien said. “We talk to our clients about ‘picking their spot,’ where it’s valuable for them to develop internal capabilities and in other areas rather than outsource to vendors.”

Even with vendor templates or tools, you still need to know what you’re doing on some level, Dresner said.

“Even if the tool is very intuitive and it’s got all sorts of heuristics built in to help you build some sort of ‘experiment’ or algorithm or script, you still need to understand what it’s giving back to you. Gee, is this really significant? Is this clustering? Golly, I don’t know,'” he said. “You can make the most intuitive tool in the world, but you still need to have some fundamental knowledge of how to wield that thing.”

But Vendors Can Help Get You Started

Your existing business intelligence tools might be one way you can jump start advanced analytics. With the shortage of data scientists, it’s likely that some companies will rely on embedded analytics or do without in the short term, Dresner said.

Already, many BI vendors support advanced analytics by leveraging R, whether that means offering basic integration in the form of hooks or in some cases entire user interfaces built on top of R, a programming language used to statistically explore data sets. “It’s a lot cheaper than trying to create your own library or having to license a third-party library to incorporate that capability,” he said. “R is just completely available and free.”

Likewise, there are startups that embed analytics into their products or services. GE Predict essentially sells analytics as a part of its maintenance offering, O’Brien said.

Look for Specific Advanced Analytics Use Cases

As with other technology investments, if you want to succeed with advanced analytics, you need to start small and stay focused on business need.

“The companies that are succeeding, instead of moving directly to prescriptive analytics that will automate an entire decision process, start with investments in analytics that will enable the core decision makers of a company to be able to make better decisions,” said O’ Brien.

In retail, for instance, rather than trying to automate the entire buying process, a company might use analytics to provide decision support tools so merchandisers and buyers better understand how customers use various products.

Digital transformation is an area that seems to be a natural fit for advanced analytics.

“I would say with respect to analytics, we are seeing our clients’ investments in digital as a strong catalyst for moving more into advanced analytics,” O’Brien said. “We’re seeing companies will invest first in getting the right digital infrastructure in place, and then recognize the power of all the data that’s coming from that infrastructure. It then becomes a catalyst to drive these advanced analytics investments and use cases.”

Advanced analytics supports both the data and the agility digital enterprises need to respond to opportunities.

“A digital enterprise is fundamentally going to be about analytics and being able to move more quickly on a particular trend as it emerges; recognize it, act on it and cash in on it,” Dresner said.

Loraine Lawson is a freelance writer specializing in technology and business issues, including integration, health care IT, cloud and Big Data.

More Posts By Loraine Lawson