Social Media & Web Analytics Innovation Summit 2014: Day 2 Highlights

Businesses now have more access to information on customer than ever before; the opportunity to provide real-time analysis of customer interactions is becoming increasingly important in identifying an advantage over competitors. The challenge remains to make use of this data to build and maintain a loyal client base.

Through investment in innovative web analytics practices and analysis of social media platforms, organizations are offered a unique opportunity to react to the data made available. To learn more about such opportunities, success stories, challenges and best practices, I recently attended the Social Media and Web Analytics Innovation Summit 2014 (May 1-2, 2014) at San Francisco, CA, organized by the Innovation Enterprise. The summit brought together industry leaders who have seen results from implementing an effective web analytics initiative to share their case studies and best practices. Illustrated intermittently with case studies, interactive panel sessions & deep dive discussions this summit offered solutions and insight from those working within the space.

It covered a wide variety of areas including measuring, evaluating & predicting the Social Consumer, Multi-Platform Consumer Engagement, Social Media ROI and Multi-variate consumer engagement and more. To help its readers succeed in their Analytics pursuits, KDnuggets provides concise summaries from selected talks at the summit. These concise, takeaway-oriented summaries are designed for both – people who attended the summit but would like to re-visit the key talks for a deeper understanding and people who could not attend the summit. As you go through it, for any session that you find interesting, check KDnuggets as we would soon publish exclusive interviews with some of these speakers.

Highlights from selected talks on day 1

Here are highlights from selected talks on day 2 (Fri, May 2):

Kevin Cooney, Director of Analytics at Wikia delivered a talk titled “Don’t Forget the Softer Side of Analytics”. In a quick introduction of Analytics, he stated that success is dependent upon asking the right questions about both the interests of business stakeholders and the characteristics of your data in your planning.  Defining term “Research” as the planning, collection and analysis of data that is relevant to decision making he mentioned that our purpose within Analytics is to become the best researchers possible.

Research process comprises of five steps: Define Objective, Design Research, Data Collection, Data Analysis & Interpretation, and Report.  The first step i.e. defining research objective, is not done most of the time or done so poorly in Analytics.

The most perfect research will be imperfect in a matter of seconds if you discover that your research answers the wrong research question.

When laying out research think in following terms: problem to be solved, what’s needed to be solved and your initial hypotheses. The second step of research process, Design research, is of three types: Exploratory (not done in Analytics), Descriptive (Surveys – Little use in Analytics, Databases – Frequently used) and Causal (Experiments – Analytics not as involved as needed). Discussing the third step, Data Collection, he explained scales of measurement: categorical, ordinal, interval, ratio. He explained the “NOIR” hierarchy as following:


Data Analysis and Interpretation is the step where we (data analysts) often start research.  He concluded the talk claiming that

tests for statistical significance (used to estimate the probability that a relationship observed really exists) do not assure that the research has been carefully designed and executed, and thus, cannot substitute logical analysis.

Itamar Herzberg, Director, Quantitative Analysis and Matt Hudson, Product Manager at Youtube shared key learning from mining the YouTube Ecosystem. They discussed three things that they have learned about data through video:

1. Choosing the Right Metric Matters A lot – Gaining Context for the metrics you measure will make them more meaningful
2. Bias is everywhere. Be cognizant of it.
3. Data Quality is always more important than Great Analysis (if one has to choose)

 
Damon Cool, Director of Analytics at Evernote talked about Evernote’s Analytics Platform. Starting with some impressive statistics about Evernote, he gave 3 laws of Data Protection as:

• Your Data is Yours
• Your Data is Protected
• Your Data is Portable

 
Giving an overview of the technical infrastructure at Evernote, he discussed various components one by one:

1. Hadoop – 10 Data Nodes with 130+ TB capacity
2. Actian Analytics Platform – Relational 150+ tables using Native columnar. This platform provides speed, data compression and deals with Ad-Hoc, Big Data and more complex queries.
3. Jaspersoft – 290+ Daily reports in various formats: HTML, PDF, CSV, Excel
4. Tableau – 10+ dashboards, 20+ users, 2 windows server

 
He then shared following key insights:

• Desktop converts better than mobile
• Desktop + Mobile converts better than Desktop alone
• Multi-platform users are more engaged and valuable but have a small share among all.

At the end, he claimed that

Analytics has changed the way the company thinks about conversion and retention.