Based on open source large data developers, data experts and IT Managers ' survey data show an interesting paradox: large data projects can provide business managers with better information to make critical decisions, but neither the business nor IT departments seem to be aware of this or are willing to pay for large data projects.
Open source Hadoop, data integration and large data analysis company Jaspersoft The survey found that 62% of enterprise internal developers are planning or actively deploying systems to provide large data analysis capabilities, whether or not these projects are marked as "large data", however, Only one-third of respondents have formal budgets or other funds to help them with these projects.
"Given the maturity of the big data market, we expect it to be a skunk project that breaks technical limits with limited time and limited resources," said Mike Boyarski, director of product marketing at the company. ”
Jaspersoft's findings may not be objective, as many of the respondents are high-tech workers.
Many respondents are preparing to provide large data analysis capabilities without dedicated large data software. They use existing analysis or database management to deal with very large, very diverse, very variable data sets.
They are also supported by forward-looking business managers who are interested in the insights that customer behavior analysis provides, but are unable to provide ROI analysis or user requirements for budget approval.
"There is a general understanding of technology among business managers, but it has not yet been fully translated into a budget, planned, clear plan," Boyarski said. "They know the importance of analyzing data of different types and different natures, but don't know what to do." So in it, technicians are trying to do large data analysis even without extra money. ”
The big data is not yet seen as a strategic asset or potential advantage, mainly because business managers do not have enough time to discover the advantages of big data, says Frank Gillett, vice president and chief analyst at Forrester.
In the coming years, the split between the advantages of big data and the willingness to pay for large data projects will continue, although the results of Forrester and other analytics firms suggest that sales of large data products will grow by 40% by 2015.
According to Theinfopro's survey, business managers are not opposed to big data, they just don't know the difference between it and traditional analysis.
The results of the Infopro show that 56% of large and medium-sized enterprises have no plans for projects involving large data after 2013 years.
Of the 607 business-unit managers surveyed by Gemini, 90% said they would make at least one different decision over the past year if they could get better information from large data analysis.
Even so, more than half of respondents said big data was not on the list that could give the company a strategic advantage.
"We have a very similar result compared to other findings," Boyarski says. "Big data projects are still like an initiative." ”
Businesses that rely heavily on the web for sales, marketing, or interaction with customers tend to be more prone to SEO, Web server log analytics, and other more professional techniques to analyze customer behavior.
Companies or business-to-business companies that focus on real-world sales cannot benefit from a deep analysis of the relatively small number of transactions.
"E-commerce companies look at customer churn and want to analyze why customers are leaving or not buying things," Boyarski says. "If they can determine the cause of the customer churn, they will be able to solve the problem more easily from the root cause." ”
Another reason, Forrester analyst Mike Gualtieri, may be the potential results from the source, nature and large data analysis.
Much of the information in a large data analysis set is still proprietary, from CRM applications, Web server logs, and other machines to machine data, rather than unstructured text from social networks, e-mails, or other sources.
Using existing tools to collect, standardize, and process these data is a huge challenge.
Almost all information management tool manufacturers are developing new large data features, attachments, or partnerships to avoid the opportunity to sell products in hot new markets. As more and more applications go into the marketplace, with more and more proprietary large data analysis and large data integration offerings, the big data situation may change.
Boyarski has a different theory: it's a matter of timing, not a tool.
"Companies trying to find customer analytics often want tools to provide them with real-time responses, in seconds or even minutes." They want to be able to respond quickly to customer needs, which will be a real challenge, "Boyarski said." Most of this data is currently in batches, and you can only make decisions on a daily or weekly basis, not hourly. ”
"As demand for real-time responses rises (especially online responses), the need for better analysis increases. We see a correlation between these two things, "Boyarski said." Many companies are still dealing with very basic issues, and less than 30% say they even use all available data to make decisions. ”
(Responsible editor: The good of the Legacy)