First, a few words about the survey itself and who responded to the survey.
- We wanted to make the survey short and sweet. We were interested in what kinds of analytical technology companies thought were important and specifically how companies were using text analytics to analyze unstructured information. Finally, since there has been a lot of buzz about analyzing social media we asked about this, as well
- Let me say up front that given the nature of our list, I would categorize most of the respondents to the survey as fairly technology savvy. In all, 61 people responded to the survey, 32% of these respondents were from high technology companies. The verticals included professional services, followed by manufacturing, financial/insurance, healthcare and pharmaceutical. There were also some responses from governmental agencies, telecommunications and energy companies. So, while the results are unscientific in terms of a random sample across all companies, they probably do reflect the intentions of potential early adopters, although not in a statistically significant manner.
- In analyzing the results, I first looked at the overall picture and then examined individual verticals as well as filtered the results by other attributes (such as those using text analytics vs. those not using the technology) to get a feel for what these companies were thinking about and whether one group was different from another. These subgroups are of course, quite small and the results should be viewed accordingly.
We first asked all of the respondents to rate a number of technologies in terms of importance to their companies. Figure 1 shows the results. Overall, most of these technologies were at least somewhat important to this technology savvy group, with query and reporting leading the pack. This isn’t surprising. Interestingly, OLAP data cubes appeared to be the least important analytical technology – at least with this group of respondents. Other technologies, such as performance management, predictive modeling, and visualization ranked fairly high, as well. Again not surprisingly, text analytics ranked lower than some of the other technologies probably since it is just moving out of the early adopter stage. Some of the respondents, from smaller firms, had no idea what any of these technologies were. And, in terms of text analytics, one company commented, ” yeekes, this must be big time company kind of stuff. Way up in the clouds here, come down to earth.” They, no doubt, are still using Excel and Access for their analytical needs. Other smaller companies were very interested in “non-cube” technologies such as some of the visualization products on the market today.
What does this mean? I would speculate that it means that companies that have found value in perhaps one or more analytical technologies are more apt to try out others. Success breeds success. We’ve seen this trend in the text analytics space. When a company deploys the technology and finds success, other parts of the organization are quick to jump on board. These technology savvy companies no doubt are gaining benefits from these technologies and are willing to invest in them because they provide significant value to the company. Those that either haven’t invested in the technologies, or are caught up in one problem –say getting their data into a warehouse- haven’t had the opportunity to experience the benefits of these other technologies, for budgetary or other reasons and so they are not that important to their companies right now.
Question 2 asked which statement best describes your company’s use of text analytics? The results here indicate a surprisingly high percent – 32.7%- actually using the technology today. Again, the reason behind this is no doubt due to the fact that the companies polled are probably very technology savvy. I believe this is significantly higher than the overall adoption rate of text analytics, which Hurwitz & Associates estimated at 10-15% of large enterprises in 2007. On the other hand, almost 45% either didn’t know what the technology was or didn’t plan to use it in the next 12-18 months. It is important to note that these results were industry independent.
We asked respondents that were using text analytics or planning to use it. what they were planning to use it for. An important point is that the majority of respondents were planning to use the technology in more than one application area. Not surprisingly, Voice of the Customer and Competitive Intelligence/Brand Intelligence led the space with over 60% of the respondents cited these two areas. Next up on the application front was e-Discovery, an area that is starting to gain a lot of attention.
Let’s slice and dice this a bit….
First up, those respondents deploying text analytics. These respondents came from a range of industries. In this group, companies are using text analytics for (in order of importance) Competitive/Brand intelligence, Voice of the Customer, e-discovery, quality and early warning and fraud. The respondents also cited other areas their companies are using text analytics for including spend analysis, analyzing company metrics, product lifecycle management, and mining business rules. Analysis of social media was very important to this group.
Those who were planning to deploy text analytics in the next 12-18 months also viewed query and reporting as most important to their company. Interestingly, the top application this group is considering is Voice of the Customer followed by competitive intelligence/brand image. This tracks with the primary research study we performed last year, indicating that Voice of the Customer should continue to be a hot area for text analytics implementations for at least the next year or so. Analysis of social networks and media was of high interest to this group.
Even those who did not have plans for text analytics were still interested in social networks – indicating that companies are starting to think about using unstructured information to gain insight about customers. These companies may be considering brand monitoring companies to serve its needs, or perhaps haven’t even gotten that far in the thought process, yet.
- Broadly speaking, companies that believe that innovative analysis technologies such as predictive modeling, text analytics, and even performance management are very important, generally rated all analytical technologies as having a higher value than those companies that did not believe they were very important.
- Text analytics continues to grow in importance, at least with this group of respondents. Companies deploying the technology today are doing so across several application areas. Companies that plan to deploy the technology are also looking at several application areas. Voice of the Customer will continue to be a hot area for the foreseeable future. E-discovery is also making noise.
- Companies across the board expressed interest in making use of social media and social networks.
Filed under: BI innovation, Business Intelligence, data analysis, Data visualization, Information Management, Text Analytics, Uncategorized, Visual Analytics Tagged: | BI, Fern Halper, Hurwitz & Assoicates, innovations, performance management, predictive modeling, survey, Text Analytics, text mining, visualization