
SOCIAL BUSINESS INTELLIGENCE IN CHINA
Abstract
OVERVIEW
INTEGRATED CRM ARCHITECTURE
Word-of-Mouth (WOM) / Social Business Intelligence is a business process that is fully integrated into Eastern Wind’s overall CRM architecture. It includes:
- Automated data collection and semantic analysis of all social conversations related to your company’s products and services.
- Insights gleaned from that analysis applied to all touch-points in the end-to-end customer life cycle.
Social media data collection and analysis is part of a broader strategy to understand prospect and customer behaviors and attitudes in the awareness, prospect nurturing and ownership phases of the customer journey. With an integrated CRM architecture, we can combine social sentiment with digital advertising and direct marketing responses, inbound emails and calls, sales contacts, requests for technical support and other points in our customer life cycle touch-point roadmap (see Digital Services – CRM).
The response to negative social sentiment may be focused on the source social media where it was identified (e.g. a forum or blog), but with an integrated CRM architecture, you can implement a broader cross-channel strategy that leverages all communications channels, including (where appropriate) messaging on your corporate web site, advertising, direct marketing, other social properties, sales collateral, etc.
INTERNET GROWTH IN CHINA
By virtue of its size, China was always destined to quickly become the world’s largest social media market. By 2016, China’s Internet user population had exceeded 700M and nearly 650M users now regularly use one or more social platforms. Since 2012, the majority of those users have accessed the Internet via mobile and an increasing number own smart phones.
SOCIAL MEDIA USE IN CHINA
CHINA IS UNIQUE
The relationship that Chinese have with social media is quite unique – for both cultural and historical reasons, most Chinese rely on social media as an interpersonal communication tool and as a primary reference for consumer buying decisions. Advertising claims are viewed with a high degree of skepticism, so Chinese consumers’ use of social media for purchasing decisions is much higher than their Western counterparts.
With the exception of LinkedIn, most Western social media platforms are blocked in China (at least periodically). So, social media platforms are highly localized – with the exception of We Chat (Weixin), most Chinese social media platforms are not widely used in Western countries. But that doesn’t mean that social media is behind the curve in China – in fact, in a number of cases, Chinese social media has outpaced Western social media, both in functionality and user adoption.
WE CHAT
The killer app in China is We Chat (Weixin in Chinese), which is a pub-sub social media platform – users subscribe to, and interact with, feeds from other users, as well as corporate accounts. We Chat has evolved from a basic personal messaging app to a sophisticated platform for marketing campaigns, loyalty programs and e-commerce.
KOL's IN CHINA
Finally, Key Opinion Leaders (KOL’s) are perceived to be major market influencers, if for no other reason than their fan bases tend to dwarf those of companies that are in still in the brand-building process. China has several classes of KOL’s, of which the most important are celebrities, bloggers and subject matter experts (SME’s). Almost all will retweet content on a fee basis.
However, the intransigent “zombie fan” problem must be considered when evaluating a marketing relationship with Chinese KOL’s. While publicly-traded companies, such as Weibo, are making significant strides in fan count verification, we typically discount fan base counts by up to 50% for independent KOL’s.
1ST GENERATION SOCIAL BI – WOM
DATA COLLECTION
First generation data collection models had limited linguistic rules engines to determine market sentiment. This has been compounded by the complexity and size of the Chinese language, which does not mark word boundaries and is comprised of over 8,000 characters, usually combined to form words of two to four characters. In addition, Chinese is highly idiomatic, with more than 20,000 four-character expressions called “chengyu”.
Initial data collection web crawlers supported a limited number of social media platforms, such as BBS and news sites, and were then extended to support blogs, micro blogs, video sharing sites, Q&A forums, SNS, and e-commerce sites. Each of these social media types has its own unique challenges with respect to accessing content, linking original content to follow-up comments and aggregating sentiment. For example, in an ongoing dialogue on a forum, the intended sentiment of one person’s agreement or disagreement with a previous mention has been quite difficult to interpret.
MENTION AND SENTIMENT ANALYSIS
First generation mention and sentiment analysis leveraged basic semantic clustering technologies to evaluate comments utilizing industry taxonomies.
However, because of the above-mentioned complexities in analyzing Chinese semantics, first generation toolsets had significant limitations in analyzing nuanced market sentiment – Chinese Netizens are particularly adept at cloaking the meaning of their opinions, especially if they are discussing a sensitive topic.
So, in practice, many local WOM analysis vendors and agencies in China initially relied on a rather crude analysis of Baidu search patterns, using large teams to manually assess market sentiment. Local vendors have often used Baidu as a primary reference and then used WOM tools to provide statistical “proof” of their initial findings.
Further, since agencies are also providing branding and campaign recommendations, root cause analysis of negative sentiment is frequently the real value proposition of a WOM analysis service offering.
DATA VISUALIZATION
First generation WOM data visualization tools have steadily improved and now include the following capabilities:
- Aggregate mentions by media property.
- Basic drill-down to specific mentions.
- Extensive filters to enhance data quality (e.g. fan verification, KOL’s).
- Automated categorization of social mentions (positive, negative).
- Retweeting reach.
- Share of voice relative to competitors.
- Search capabilities within the collected data set.
However, meaningful analysis remains bound by both limitations with semantic clustering techniques and the above-mentioned limitations of Chinese language data collection and semantic interpretation.
2ND GENERATION SOCIAL BI – NLP
LANGUAGE INDEPENDENT FRAMEWORK
Second generation WOM data collection, market sentiment analysis and data visualization reflect a global approach to Social Business Intelligence – that is, they are based on a language independent framework that allows for a common rules engine for core processes and individual natural language modules to handle the unique requirements of each target language.
The benefits of a common framework for global Social Business Intelligence capabilities are not lost on both local Chinese companies and MNC’s who need a common analytical capability for global markets, as well as an effective tool to decipher local sentiment in Chinese social media.
SENTIMENT AND TREND IDENTIFICATION
Second generation Social Business Intelligence technologies, such as Synthesio’s Automatic Sentiment Analysis (ASA), use a combination of machine learning, sentiment dictionaries and linguistic rules engines to identify “opinionated text blocks” (OTB), which are further refined with weighting rules to modify and in some cases, reverse) the sentiment value of an OTB.
In the case of Chinese, a machine learning rules engine iteratively learns to recognize word patterns in the context of sentences (as noted above, Chinese sentences do not use word separators).
As previously noted, correctly identifying the context of a social mention is an extremely complex (and sometimes subjective) process. Second generation tools continue to automate context detection using machine learning techniques, so that irrelevant mentions are filtered out during data collection.
Finally, second generation Social Business Intelligence tools have moved from relatively primitive positive / negative aggregation to enhanced topic mapping capabilities. Administrators can “seed” the machine learning process by manually assigning topics to a limited set of mentions and the tools apply those patterns to subsequent analysis.
EASTERN WIND’S SOCIAL BI PRACTICE
STRATEGY
While Eastern Wind focuses primarily on the Chinese market, we work with clients to identify short- and long-term social sentiment tracking requirements that reflect expansion in APAC and Western markets. Not all industry sectors benefit equally from Social Business Intelligence programs – factors that influence the usefulness and timing of Social Business Intelligence implementations include:
- Markets served – China, APAC or global.
- B2C vs. B2B distribution models.
- Current market awareness.
- Competitor market awareness.
- Planned product launches and marketing campaigns.
- Exposure to reputation crises.
- Capability and maturity of the organization’s end-to-end CRM processes (see below).
PLATFORM SELECTION
From a practical perspective, our experience with Chinese social media and Chinese language sentiment analysis suggests that clients consider the trade-off of platforms written specifically for Chinese (CIC is a local leader) and platforms that offer enterprise scalability across multiple geographies and social platforms and the potential to bring sophisticated semantic analysis techniques to China (Brandwatch, Synthesio, Crimson Hexagon and others lead in this category).
CRM INTEGRATION
Since digital marketing campaigns directly (and indirectly) integrate social platforms as communication channels, Social Business Intelligence is increasingly an integral component of campaign management, serving as a critical feedback mechanism that supplements direct campaign response metrics.
By integrating social media and Social Business Intelligence in branding campaigns, organizations can analyze events and campaigns in near real time.
By integrating Social Business Intelligence in loyalty programs, organizations can respond to real consumer concerns in a meaningful way, reinforcing positive feedback and responsibly addressing negative feedback before it becomes a crisis.
In the event of a crisis, Social Business Intelligence can provide real time monitoring of the effectiveness of any measures used to protect your company’s reputation.
At all points in the customer journey, Social Business Intelligence can substantially increase the ROI on marketing and CRM spend.
ABOUT EASTERN WIND
Eastern Wind is a digital consultancy based in Shanghai, PRC, providing local Chinese and multi-national companies with digital strategy, program management, CRM and digital marketing systems design and integration services in the financial, auto, pharmaceutical, medical devices, luxury goods and FMCG sectors.
Year of the Pig

Offices
-
US: Manchester, VT
CH: Shanghai, PRC - US: +01 415.870.0962
- CH: +86 950.135.100.0557
- [email protected]
Insights X Industry
Insights X Topic
.