Chinalytics Longmarch Blog

Yesterday I had the pleasure to host the July WAW here in Beijing with John Wang, Product Manager at Sinotech Group sharing Sinotech’s perspective on Web Analytics and Social Media Analytics. Thanks John for sharing.

While John had only a couple of slides, it was slide 8 (the SinoBuzz dashboard) that caught most attention and started a discussion that lasted almost half an hour. Thanks John for being a great sport answering all the questions in detail (especially the more pesky ones from Paul Denlinger )

Some take aways from the presentation and the discussion:

• The Social Media tracking field is getting crowded, with pioneer CIC, BBS aggregator Daqi, Ogilvy One’s OBuzz and now Sinotech’s SinoBuzz. That’s great news for advertisers and ad agencies. It keeps everyone on their toes
• Sinotech has been working of Sinobuzz for more than one year. Especially the natural language processor part had been tricky, but they are confident that it produces reliable results now.
  FP: Chinese is a very difficult language to interpret with software. None of the natural language processing tools I have seen for Chinese have worked at an acceptable level. I would love to see a more detailed demonstration of Sinotech’s system on a post and aggregated level to understand if they have cracked this particular nut.
• SinoBuzz is working with connectors to plug into BBS, Blog, Social Network, Video sharing and other conversations. The platform is flexible to add other connectors to any upcoming platform or network.
  FP: This flexibility is a clear competitive advantage, given the more static nature of most competitive products.
• SinoBuzz is trying to aggregate unique metrics from various platform
  • #of BBS post, BBS clicks, BBS replies, Post Sentiment
  • #of Blog post, Blog replies, estimates about Blog post PV’s, Blog post sentiment
  • SNS #of Friends, #of status message, #of posts, # of relies, post sentiment
  • etc

  into general metrics like “reputation” and “attitude”.
  FP: I am highly sceptical of this approach since is a) lack transparency b) ignores unique characteristics of different platforms, their audiences and their impact c) uses unreliable third party resources like Alexa to gauge the impact of sources that do not publish reach (PV) data.

• The Sinobuzz dashboard is well designed and provides a good overview of brand performance. Like most dashboards (see Avinash’s Action Dashboard post), it suffers from the lack of analysis and actionable insights.
  FP: A good rule of thumb for any Dashboard: Handle with care. Dashboards without analytics and recommendations usually end up as pretty charts without impact. The key resource is not the dashboard, but an analyst who understands these numbers and your business and can provide actionable insights and recommendations based on that data. John mentioned that there are more reports to dig deeper and it would be great to understand what these reports can do. The critical queastion remains: How does the tool enable brands to take action?

What are your thoughts? Did I miss anything important? Let me know in the comments section.

Related articles by Zemanta

• Defining Social with Usage Data (arnoldit.com)
• Semantic Search Getting Social (marketingpilgrim.com)
• Accidental Measurement And Why We Love Useless Metrics (rohitbhargava.typepad.com)

Daqi Logo

As a web analysts, we try to understand and optimize the performance of our or our clients’ websites and digital campaigns. For the most part that has included media optimization, creative opmtimization and site optimization. In a Web 2.0 world that is not enough anymore. As brands realize that they share the ownership of their brand image with their customers, prospects  and other interested parties online, (who influence each other by sharing their impressions, opinions and feedback trough ever increasing social media channels), brands are asking us to help them make sense of the discussion happening about their brand online a s well.

Zhou Lei will introduce Daqi’s tools and services that enable brands and agencies to keep track of the  ever proliferating online discussion. She will share some examples from auto/IT/FMCG case studies and describe  Daqi’s methodology. A quick intro to their service:

Daqi’s Buzz Radar system monitors 700,000+ BBS and 10 blog service providers in China. We capture av. 500,000+ data points from social media sites every day. From 2004 to now, our database has 8000+ million data entries. Based on the primary data from social media, we provide market intelligence, ads/campaign tracking, industry analysis, online behavior analysis, social media analysis,  IWOM analysis reports for our clients.

Please join Zhou Lei, me and more than 30 other web analytics enthusiasts to learn about web analytics, meet other web enthusiasts and have an all around great time. Bring any friends who might be interested to join our community along as well to:

Location: Luga’s Villa (right behind 3.3 in Sanlitun)

Time and date: Wednesday April 1st , 8PM

We will have a buffet dinner and soft drinks available for our guests. Be prepared to spend RMB 50 for the evening. As usual the knowledge you get in exchange is invaluable

Related articles by Zemanta

• March Web Analytics Wednesday Beijing with Google’s Stephanie Hsu (longmarch.chinalytics.com)
• Using Web Analytics to Meet Your Business Objectives (edocr.com)

As a nice follow up to my post on online metrics in the world of user generate content, I had the chance to have lunch with Sam Flemming of CIC / IWOM fame. While I have been somewhat critical of CIC’s reports, I do admire Sam as a visionary and Chinese internet pioneer. CIC almost single-handedly popularized Social Media marketing and measurement (they call it Internet Word of Mouth or IWOM) in China, a market that
has not even started to measure online activity on company websites effectively.
Most of our discussion focussed on how to integrate social media measurement, an area where CIC has patent pending technology, with web analytics data, CIC’s positioning towards agencies, using social media effectively and the Chinese Social Media environment in general. But one thing after the other (or Step by Step, as NKOTB said, yes I still can sing that song)

Chinese Social Media environment

• China has a very vibrant social media scene that is driven by BBS culture
• Social networks are growing out of successful BBS’ and this history makes them unique and uniquely different form popular western networks
• These networks are populated mainly by young netizens, high school and university students
• The discussions hosted on these networks cover a wide range of topics, many of them relevant to major consumer brands
• Currently there is no equivalent content rich network for business users (while there are LinkedIn clones they do not feature discussions), although companies like Alibaba and Xing are moving into this space

CIC’s positioning towards agencies

• Sam sees online agencies as partners for CIC and has positioned the company as a compliment. They will measure social media conversation and consult clients on a IWOM strategy, but it will not execute or manage IWOM campaigns or seed posts.
• In reality both agencies and clients don’t see this distinction as clearly and advertisers do position CIC and agencies as competitors. More communication is needed to create a win-win situation. Our  discussion was good step into this direction.

Effective usage of social media

• Once marketers understand that their brand is talked about online and how this influences their brand equity, their natural response is asking “how can I influence the ongoing discussion”. While this is a valid question, the easy answer (seed your own posts or hire a 3rd party vendor to control the debate) raises serious ethical issues and will, in the long run, destroy the trust and effectiveness of internet word of mouth. Especially multinationals should be aware of the risk to their reputation that comes along with faking posts.
• Sam believes that this traditional media approach (just shout larger than everyone else to make your message heard) is too simplistic and recommends advertisers to listen first then be understood. (Wise words if you ask me) Understanding the discussion going on online can inform advertising campaigns, since it works like a giant focus groups (free of charge). In a next step advertisers are also able to get almost real time feedback about users take aways from ongoing marketing efforts, and can adjust their campaigns to better resonate with their audience.
• IWOM tracking can also be an early warning system. Brands can learn about problems with their products, communication or distribution by listening to the unfiltered voice of the customer
• For brands to participate in the discussion (note the difference to ‘control the discussion’) Sam recommends a more sincere approach that centers around ‘real people’ sharing their personal opinion in a casual language.
  He notes the example of Dell, a company that effectively used the ‘personal’ blog of senior managers to enter into an open discussion with users that previously voiced their frustration on sites like Dell
  Hell. An interesting question will be how to translate such blog based effort into the BBS culture in China (Any ideas? Let me know in the comment section). I do strongly agree with Sam that in the
  long run only sincere and real participation will be rewarded. But China being China, many a company will look for quick fixes that will in the end harm their brand.

Integration of social media measurement & web analytics data

• To enable a better integration of IWOM data into regular web analytics reports, along the lines of Dennis Mortensen’s Online Business Media Quadrant Model, I highlighted some challenges of CIC’s current offering to Sam

1. Timeline: My main gripe with CIC’s “CIC data” report is timing. It usually covers a 1 month time frame, while ad campaigns usually start sometime in the middle of the month and end 4,6,8 or 12 weeks later.
   Aligning these timelines is almost impossible. In addition CIC’s report arrive with 2 weeks+ time delay after the month is over (most of this, I am sure, due to the insight mining that takes place), that is too long for kind of near time reporting I am looking for. Sam assured me that their “CIC Alert” product provides more flexibility in terms of timing (daily, weekly or monthly) and can be adjusted to follow the campaign duration. This type of reporting would integrate with Web Analytics efforts more effectively. To take that one step further we talked about access.
2. Access: Currently CIC delivers its reports in PDF formats by email. I would love to have a web interface, much like my web analytics tools, or IResearch’s IUserTracker (link in Chinese). That way I could download the quantitative info (number of posts, share of positive, negative, neutral) whenever I need it (I can wait for the qualitative insights by PDF), manage the keywords I would like to have tracked (product names, campaign names, competitors etc.), organize them into campaigns, specify the timeline I want to analyze and then export all this data to an Excel spreadsheet and compare and correlate to my other online data. Doesn’t that sounds like a nice little IWOM vision

• While I am not at liberty to share more details, I can say that Sam was certainly listening intently and I am curious what they have in the works. Stay tuned. I certainly will.

Related articles by Zemanta

• Social Media in China
• The internet is the community – Sam Flemming
• “Group purchase” – ecommerce with Chinese characteristics
• Social media more important than search engine optimization for lawyers