Text Analytics and Product Management

March 27, 2010

In the last post I looked into how text mining techniques are becoming more important for customer experience management, customer relationship management and in creating competitive advantage in marketing through understanding the customer sentiment. As noted, there are many software companies that are actively targeting this sector and offering increasingly sophisticated tools for the business users.

In this second post on text mining, I want to look further ahead, and see what are the opportunities for text mining technology in the future, and how it can be extremely useful to the product lifecycle management solutions.

While today’s text mining applications are able to handle a large volume of unstructured data gleaned from multiple sources, they are yet able to do so in real time and provide a flexible enough interface for users to manipulate the data as they come in. As the development of network traffic logging tools has shown, what is key to bringing these tools to the next level is the ability to deal with ever larger amount of data in less time and with more flexibility. Instead of offering cookie cutter analysis or hard-coded analytical templates, text mining vendors should think about building modular components that can be mashed up together easily by the users to create unique analytical engines used for particular content domain or content type.

Moreover, as another constraint on the speed of processing is obviously the amount of computer horsepower available to the application. The text mining platform should be smart enough to automatically balance the resources and bandwidth afforded to its different components – such as the data logging database, the semantic analysis component, the statistical analysis back end and the charting engine,

Text mining, as applied to customer feedback management, should be able to handle the nuanced differences between data sources and data formats, for example, the different reliability of directly recorded feedback versus feedback gleaned indirectly from the Internet.

Lastly, I still think that text analytics are underused inputs into the development of products, as through text mining, the essence of the customer experience is surfaced and brought directly into the product management process. However, because text mining was not originally built for that purpose, its integration with leading product management solutions is spotty at best. I would like to see more specialized version of the popular text mining packages built for product manager or product marketers.

Continuing in this train of thoughts, text mining can be offered as part of a toolkit for product manager, putting it in the same categories as other product management tools such as the Innovation games etc. It can be even more tightly integrated with those tools if the transfer of input data and outputs are simplified through standards-driven formats and interoperability.

To close off this post, here is a list of text mining resources, toolkits and relevant sites for reference.

Chief Business Officer at UserTesting

Tien Anh joined UserTesting in 2015 after extensive financial and strategic experiences at OpenView, where he was an investor and advisor to a global portfolio of fast-growing enterprise SaaS companies. Until 2021, he led the Finance, IT, and Business Intelligence team as CFO of UserTesting. He currently leads initiatives for long term growth investments as Chief Business Officer at UserTesting.