IT-Analysis.com
IT-Analysis.com Logo
Business Issues Channels Enterprise Services SME Technology
Module Header
Craig WentworthMWD Advisors
Craig Wentworth
16th April - Egnyte the blue touchpaper...
Louella FernandesLouella Fernandes
Louella Fernandes
11th April - Managed Print Services: Are SMBs Ready?
Louella FernandesLouella Fernandes
Louella Fernandes
11th April - The Managed Print Services (MPS) Opportunity for SMBs
Simon HollowayThe Holloway Angle
Simon Holloway
11th April - Intellinote - capture anything!
David NorfolkThe Norfolk Punt
David Norfolk
11th April - On the road to Morocco

Blogs > MWD Advisors

Cognos Consumer Insight: blending social media data with predictive analytics
Helena Schwenk By: Helena Schwenk, Principal Analyst, MWD Advisors
Published: 8th June 2011
This work is licensed under a Creative Commons License
Logo for MWD Advisors

At its recent Business Analytics Summit, IBM dedicated a session stream to its social media analytics platform Cognos Consumer Insight, an offering that allows organisations to gain a deeper view on consumer attitudes and brand preferences. Given the hype and resulting confusion currently surrounding social media and its value to marketers this was a useful and opportune moment to take a closer look at the offering.

Consumer Insight is designed to help marketing professionals maximise the value from social media data by getting a firmer grip on its content, distribution and influence. The offering works by automatically identifying and tagging relevant content across a range of social media sources including blogs, twitter, news feed and review sites based on queries that search for specific words or phrases. The data or snippets collected in the search result are loaded into a database and made available for analysis. Through its integration with Cognos BI the results can be served up in predefined dashboards providing information on coverage, hot words, influencers and sentiment analysis. The latter capability allows users to determine whether consumer comments are positive, negative, neutral or ambivalent. Likewise advanced features such as affinity analysis allow users to detect relationships between hot words, whereas an evolving topic analysis allows users to follow trends and common discussion topics across time and related keywords.

The market is crowded with vendors offering media monitoring or listening solutions. As such IBM aims to differentiate itself by providing a more scalable and analytically advanced solution by tapping into some of the assets within its vast Business Analytics and Information Management portfolio. In particular Consumer Insight leverages the capabilities of SPSS dictionary and sentiment rules, as well as InfoSphere Big Insights and Hadoop, to allow Consumer Insight to process the billion plus types of unstructured social media content on the web. While IBM is marketing Consumer Insight as an analytic application, not all of the content and components are necessarily pre-packaged. For example users still need to invest time and effort augmenting or refining the text dictionary to configure and customise it according to business and industry needs; likewise fine tuning the dictionary on an ongoing basis is important to ensure it re-learns as it goes along.

That said, it is still early days for Consumer Insight and the social media analytics market in general. So far IBM has found most of its success (not surprisingly) within the Consumer Packaged Goods industry. However the company is looking to broaden and extend its reach to other industries and sectors. One of the key challenges governing its uptake for IBM (as well as other social media vendors too) is to help companies determine how they define, categorise and measure the value resulting from their social media efforts and investment. One of the ways IBM intends to tackle this issue and outpace the competition is by integrating its social media and predictive modelling capabilities in a way that helps drive and enhance marketing processes such as those for customer retention, building advocacy and loyalty.

From a technical perspective this means deepening the ties between Consumer Insight and SPSS, allowing social media data to be dropped into SPSS Modeller so it can enhance and augment predictive models. There are a number of use cases where this combination of social media and predictive modelling could become particularly pertinent. For example, in the area of predicting churn, insights from social media (such as negative sentiment about network coverage for a particular Telco) could be used to help predict those customers who are likely to defect. Similarly, understanding the key factors driving sentiment across product or brands could allow organisations to build predictive models that increase positive sentiment. Equally, one of the newer use cases involves using predictive models based on the analysis of social media data to identify the propensity of someone becoming a customer advocate. These insights could in turn be used to run online and marketing campaigns (via Unica) to leverage their influence in certain high value communities.

We believe these examples are helpful in building awareness of use cases and bring greater clarity into how organisations can derive actionable insights from social media data through better segmentation, campaign targeting and predictive capabilities. The challenge for IBM will be to build on this awareness, build referencable customer case studies and educate the market on how social media analytics can really make a difference to revenue growth and marketing ROI. We look forward to reviewing the progress Cognos Consumer Insights makes in delivering on these aims.

Advertisement



Published by: IT Analysis Communications Ltd.
T: +44 (0)190 888 0760 | F: +44 (0)190 888 0761
Email: