What is Social Data?
Social Data seems to be the new buzz word in the social media sphere, but it often ends with simply figuring out whose digital voice is the loudest in the market. As with everything else in digital marketing, it is not about the quantity, it is about the quality combined with intelligence (in an ideal scenario). Social media is no longer a ‘read and respond’ tool, it has now become a valuable channel for brand reputation measurement and management, customer acquisition and engagement, and hopefully an attributable ROI platform in the near future. Social media often becomes the first point of contact with a brand or product, but it is not your brand page consumers reach out to in a first place. They rather seek friends’ recommendations, expert advice or product reviews. You have to actively observe and interact outside of your comfort zone, which is your owned and controlled brand page. In order to do this effectively, you have to build a strong social data management structure which eventually should become the foundation of your brand’s social strategy. If you have a social media strategy without social data infrastructure, it means you have no strategy as yet and hopefully this article will be your guide to building a successful brand online.
Consumers own your brand reputation, but you can influence it
Social media users are the most active product reviewers online, as they actively share their feedback, provide recommendations, praise and criticize brands. This trait has given them the power to create and own brands online. It is no longer about what brands say they are, it is about what users say the brands are. According to Joe Tripodi of Coca Cola, over 80% of Coca Cola related content views online are from consumer-generated content. Social media is the most powerful form of WOM (word of mouth), and the best way for companies to control it is to keep their product quality and customer satisfaction high. Social media analytics is one of the tools to help you get there, but only if you do it right.
Loudness doesn’t matter
If you are an active social brand, most probably you have millions of tweets, thousands of Facebook likes, hundreds of Foursquare check-ins, and another heavy load of newly added Pinterest and Instagram data. We are at the doorstep of social data explosion which will soon become completely out of control. What are we going to do with all of this data, where are we going to store it and how will we act on it? We already struggle with the existing data pulled from loyalty and CRM programmes, customer transactions, personal information obtained from account set-ups, web analytics, etc. All of this data tends to have clear structure and data mining systems in place, unlike social media which is comprised of a huge bulk of disorganized information with no assigned values or context.
Social data can’t be isolated
Looking at your social channel statistics (shares, re-tweets, likes, comments, etc) is a very abstract view which is hard to analyse and gives no direction or any insights to help you build your brand online. How many of your social fans visited and interacted on your website, how many of them saw your TV ad, downloaded your mobile app and called your customer service? To see a bigger picture you need to draw correlations, user paths and behavioral11/28/2012 changes. Most of the time you will find your fanbase not to be your target market. This creates a skewed understanding of your existing and potential consumers. Measuring engagement rates won’t give you a clear picture either, as they fail to identify actual buying intentions. If someone visits price comparison websites for iPhone, how does it relate to someone reading or commenting on a post? Social data cannot be measured in isolation, other analytical tools and channels have to be integrated in order to identify a social user’s purchasing funnel, their past behavior and intent to purchase your products. Data integration can become one of your most powerful tools and help you understand what triggers certain actions and what behaviors demonstrate intent.
Where are we with social media analytics?
While some companies are just starting to realise the importance of social media monitoring and are making their first steps in data gathering and analysis, others are way ahead syndicating their social media data with predictive processes. According to Gartner’s senior research analyst Jenny Sussin, the overall target market penetration for social analytics is in 5-10% range, and in terms of companies doing something predictive it’s less than 5%. However a recent survey by Hypatia Research Group shows an increased interest in this area, with nearly 60% of respondents planning to allocate 1-2.9% of their annual marketing budget to social analytics next year.
The key approaches to social media monitoring:
Reactive: Channel Reporting (beginner)
Limited to gathering and analyzing specific channel (i.e. Facebook, Twitter, Youtube, etc) data, this approach is good for social media beginners as it gives an overview of a certain activity, helps to better understand channel characteristics and social behaviors, it is also a good tool for response management. However this type of data is not useful for gaining deeper insights into your fanbase, their behavior, trends and customer lifecycles.
Descriptive: Social Media Monitoring (intermediate)
These cloud-based tools crawl and capture information from various social media channels and provide summarized data in graphical or tabular formats. These tools provide a good overview and perspective on specific topics, what consumers are talking about your brand, post popularity and brand share of voice. The disadvantage is that the data cannot be enhanced or contextually analysed. Besides the providers often overlook niche sites (industry specific) or exclude certain sites from their package.
Predictive Analytics: text-mining (advanced)
This technique performs sentiment analysis and generates fact-based insights by translating text into various contexts, such as positive and negative. The success of this method highly depends on the lexicon set up (what is perceived positive, neutral or negative context), the more advanced the lexica the more accurate and insightful data you can get. Example: ‘Product X is extremely good and useful, but a little bit expensive’. This review has two positive words (good, useful) and one negative (expensive) therefore it would be tracked as ‘positive’, especially because positive word is enhanced by ‘extremely’ while the negative is minimized by ‘little bit’. The disadvantage of this technique is that it doesn’t provide the information about the creator of the message, its visibility, the numbers of readers and responders, and community reaction.
Predictive Analytics: network-mining (advanced)
This technique identifies how individuals interact with each other and does not rely on ‘likes’ or star ratings to identify the importance of a person. It rather identifies key influencers and followers based on physical nodes and connectors. This method is very good at identifying glitches such as ‘I’ll vote for you if you will vote for me’ which cannot be solved by text-mining. Besides the network analysis, this methodology lacks contextual data.
Predictive: Hybrid of text-mining and network-mining
This is the most effective way to identify and interact with the most influential and impactful users who have shown a positive or negative attitude towards your brand. It also allows you to create groups of individuals with similar characteristics and interests, such as products they use, social activities they engage in, etc.
Can’t be alive without Social Plug-Ins
Aligning social media data with CRM systems is one of the most promising data validation methods, yet it is the biggest challenge we face nowadays as it requires manual data entries and integration. Whereas social sign-in and mobile open authentication are the new trends which allow users to sign into restricted sites using existing sign-in data instead of having to create a new account. Besides creating a simple and easy user experience, it provides an integration platform between your website analytics and social data. It gives you deeper insights into user preferences, behavior and purchasing funnels. And remember, your website is your priority # 1. Instead of pushing Facebook traffic you want your social media channels to become key traffic drivers to your website.
Ask, but make sure you are asking the right person
It is anticipated that one multi-branded hotel company will replace its traditional survey and use social media as their main customer measure of experience. If you are using social surveys, forums or discussion boards to help you design new products, make sure you are connecting with the right target audience and have a strategy in place to collect, analyse and utilize this data. Otherwise you’ll end up with data which is stored but not used, and disappointed customers who took the time to participate but didn’t see any outcome.
Social apps don’t fit all
Social apps is another trend which quickly turned into a mass marketing activity. Remember that not all of your fans are the same; they all have their individual preferences, needs and interests. Utilise your social data to personalize your apps and make them more relevant to a specific user. Walmart launched a Facebook app called Shopycat which makes gift recommendations based on friends’ Facebook activities, including their posts, shares, and likes. It matches friends’ interests to Walmart’s catalogue, as well as external gift shopping and gift card sites, and provides tailor-made recommendations and store locations to collect the gifts. This example demonstrates how data analysis can enhance social media apps, improve their relevance and add value to the consumer.
Social media marketing gave us the power of speech, but we often end up speaking in an empty auditorium. Social media analytics gave us the wisdom, but we still don’t know how to use it. Ideas and science are the key drivers of innovation, so be creative and use social data to implement great ideas. Target wisely and always be relevant to the user. Sending a honeymoon package offer to a fan whose relationship status changed to ‘engaged’ is a good example. Acknowledge your brand advocates and reward your loyal customers.