Social media analytics for noobs

This article is meant for people like me – newbies in the world of social media analytics. I suspect a lot of people working on web analytics would also feel like beginners since this field is still in its inception. A lot of areas are still unexplored and social media has the tendency to consistently define and redefine itself as it grows. Below are some standard concepts on social media analytics and my take on it.

Overview
Since the Internet acquired a more public face with the inception of the World Wide Web, a greater part of the 90s saw industries trying to understand this new medium and answer the question, “How does it work?”. In the early 90s, a lot of experimentation and trials were conducted attempting to effectively leverage the answer to the question ‘What can we do and get from it?” Amidst all the hustle and bustle that accompanied this new medium, business have now started to turn their attention to knowing how to measure the effects of this medium and harvest its insights. How to spin the proverbial hay into gold?

The importance of web analytics — the measurement and analysis of user intent and behaviour online (what they say and what they do) and transformation into what Avinash Kaushik calls ‘actionable insights’ — to the success of online projects is now recognised by major businesses as evidenced by its growing role in the product / service life cycle.

Internet technologies actually enable the medium to be more measurable compared to others, such as TV, and thus can contribute more to accountability and transparency. This even becomes more relevant in times of financial crisis. The growing number of free web analytics technologies is helping more and more businesses to test the waters and discover how analytics can best serve their goals. At the same time, a shift in focus is occurring from mere data generation towards more analysis to dig out actionable insights that can be fed back to brand /product research and development, marketing, sales and post-sales activities.

Enter social media
Availability of and accessibility to more bandwidth has contributed to increased Internet access and diversified usage. It has become easier to share, produce, store and distribute content. The arrival of social networks provided users a same-but-different kind of platform to communicate and connect with each other. It’s a phenomenon where technological platforms and social dynamics feed into each other resulting into various social online channels, which fuel different kinds of conversations. Companies are trying to make their presence heard and felt within this ecosystem. As they increasingly devote more resources in building their social media presence and identity, questions on this media’s effectiveness are raised. It may operate digitally, but it is not as black and white as clicking on a shopping basket.

What is social media analytics?
It’s a way of measuring the success of your social media approach: your listening, engagement and learning strategy. Simply put, social media analytics is the analysis of online conversations.

Social conversations can revolve around your company, products, issues, competitors, and campaign outcomes. By quantifying trend, sentiment of references and influence, online conversations can be monitored, measured and analysed.

Take-away: As with web analytics, analysis requires a combination of automated systems and human insight to turn raw data into actionable insights.

Why track online conversations?
Some basic motivations:

• Know which product or campaign is making an impact and what needs to be managed.
• Know the sentiment of consumers, gain more insights into their needs and concerns and be able to address timely those concerns
• Identify the influencers, listen and learn more about them and decide on a path of engagement. How can you get them on your side?
• Feed these insights back into the product / brand life cycle.

Take-away: In the end, it’s important to distinguish the noise from the message, and chatter from the conversations in order to pinpoint the “biz from the buzz”. But to do this, first you must immerse yourself in the whole process and be a genuine part of the ecosystem.  Your tweets with another person may not result in a direct sale, but taken together, your conversations are foundations for building a relationship that you can leverage in the future.

How to measure success in social media?
Different businesses have different objectives. Strategies, tools and measurement of effectiveness are going to vary. Social media is NOT the means and end for everything. It can deliver some types of successes better (e.g. interaction vs. lead conversion; or post-sales vs. actual sales), so some more traditional metrics may not be the best success indicators.

Take-away: The question about metrics – whether for social media or not – is: what data would be valuable to you, how to measure it accurately, and if the insights it provides are actionable.

Some common social media metrics
There are different metrics that can be used for blogs, networking sites, applications or widgets.  IAB released a document specifying some standard definitions for social media ad metrics.

Some general metrics / areas for social media metrics:

• Attention / Captivation:  The amount of traffic to your channel or content, like page visits.
• Participation / Interaction rate:  Level of engagement of users who interact with an ad, application or channel. E.g. video installs, ratings, comments. Some providers use a metric called Conversation Index, which is the ratio between posts and comments + trackbacks.
• Author credibility / Authority: number of links to your content or links to other content containing conversation phrases identified in the campaign.
• Influence:  The size of the user base. E.g. blog subscriptions, application fans
• Impact: What the influencers did with your content. Did they review it, re-tweeted, blogged about it?
• Velocity: Rate of how fast your message is travelling in a given time.
• Longevity/Lifecycle: Average period of time for which an application or widget remains installed by a user.
• Content freshness and relevance: rate of posts being produced or circulated. E.g. earliest, meantime, latest post dates

Examples of actionable insights that can be fed back to the product / brand life cycle

• New customers & prospective markets
• How to reduce buying cycle & support costs by encouraging self-support
• Involvement in the product development cycle
• Increase sales / encourage loyalty through customer satisfaction derived from customer
• Increased efficiency in developing products due to customer feedback at various stages
• Minimize brand damage by responding quickly to customers’ concerns online

How companies have benefitted from analytics
Note: The first two examples are from a ppt document from Nielsen BuzzMetrics.
For the example on Dell, go here to read the full article.

Supporting sales (Toyota)

• Situation: Potentially critical post-delivery issue on best-selling Camry.
• Approach: Identified affected Camry owners “complaining” online and analyses their comments.
• Action: Proactive online communication and transparent customer service response.
• Result: “We’ve saved customers, we’ve improved products and we’ve strengthened loyalty because of this effort.” (Bruce C. Ertman, Corporate Manager, Toyota Motor Sales)

Trend tracking (ConAgra Foods)

• Situation: ConAgra, (US agribusiness conglomerate invested in health / diet foods) that cater to a recently trendy diet, receives an early warning that the diet craze is fading.
• Approach: Monitored and analyzed online consumer conversations to determine whether sentiment for the craze is negative or positive and whether craze is on way up or out. Analysis shows interest in the diet is, indeed, fading.
• Result: The company seizes the opportunity to promote alternative food products and to begin development of foods that meet consumers’ expressed desires for new product types.

Customer service (Dell)

• Situation: Growing number of complaints on customer service. A blogger’s rant on “Dell Hell” snowballs into a blog saga and was covered by tech blogs, newspapers and magazines.
• Approach / Action: Dell developed social media strategy and applied it to customer service among others.
• Utilised sentiment analysis of conversations tracked by topics to understand what the conversations are about.
• Identified concerns earlier than they would’ve had previously thanks to blog and social media commentary.
• Tracked revenue generation from The Dell outlet, small business and home offers available on Twitter. @DellOutlet (620,000+ followers) has surpassed $2 Million in sales (June 09)
• Contributed to and integrated with product development via the Ideastorm community. Generated 10,000 ideas and over 200 implemented.

Companies offering social media analytics services
There’s a growing list of companies offering social media analytics services. It seems a bigger majority focus on online aggregation of opinions and some couple this with sentiment analysis. So, it’s not just about the number of users who liked my brand, for example, but also, what exactly are they saying about it and what do they mean. A few, and relatively more expensive ones like Crimson Hexagon, attests to their “…ability to track and measure opinion according to business criteria, and to calculate ROI of online marketing and PR investment.”

Check out the following links for more resources:

• Jeremiah Owyang: Companies that Measure Social Media, Influence, and Brand
• Literally a super list for social media analytics resources
• Philip Sheldrake’s list of social web analytics vendors (over 70 and growing). He also has a free ebook on this topic: The Social Web Analytics eBook 2008

=============
Social media analytics is a very new field and a lot remains to be explored. It may be a new, emerging medium, or a new sub-set of digital media, but it’s still very much rooted in the same context and ecosystem where other media exist. While new tools or methods may be needed to quantify it better, we can also turn to existing metrics, approaches or their derivations – and common sense — to better understand how to identify and measure success in social media. Its novelty shouldn’t prevent us from us from utilising means that already work, but it should make us ask different questions. And when necessary, find new answers.