Creating and publishing data without analysing its credit-worthiness is like cooking the food and not serving it. There are no two ways about the fact that analytics is a crucial element in any marketing campaign. Apart from website analytics, social media analytics is something that enables you to get an overview on how to go about building your social media analytics dashboard. Gain insights on the same through the webinar led for Digital Vidya by Anshul Jain, Founder & CTO, ThoughtBuzz who runs a social media analytics company himself.
Q&A During Webinar With Anshul Jain
Q- What is meant by social media analytics?
A- I’ll answer this question step by step. When we talk about social media analytics, we don’t just mean Facebook and Twitter, unlike a lot of brands in India and other countries in Asia, who have been able to generate results through advertising on these platforms. Such companies do not have structured content on a 3rd platform like YouTube, Pinterest and other discussion forums. When we talk about social media analytics, we take a 360 degree perspective, which means that Facebook and twitter are a part of it but we also try to see the kind of comments people post on discussion forums and how the discussion unfolds. So, social media analytics needs to look at the entire gambit of online environment and user-generated content, irrespective of the platform. We urge clients and in fact, I would urge you to search for some of the keywords of a brand or a product or a service on such forums and platforms to see where the brand stands online from a user-sentiment perspective.
Q- How does social media analytics work?
A- Social Media Analytics requires a dashboard to begin with. Each dashboard has its own USP but primarily, most dashboards look at 3-4 components:
- Sentiment tagging: Dashboard tags the sentiment as positive, negative or neutral.
- Buzz trend analysis: Breakup of the buzz over a period. Has it peaked? Has it fallen? Is it in line with the campaign launched? So on and so forth.
- Geography: The common platforms where the discussions take place.
Any information used in social media analytics is the one available on public domain and is accessible. So, information on protected Facebook and Twitter profiles can’t be captured. Then, social media looks at ‘word clouds’ – the discussions associated with a particular keyword. Social Media Analytics includes the people who regularly talk about your brand or product. Some dashboards offer the ability for you to see who is talking about the brand or product on which platform or across which platforms.
Q- What can social media analytics do for you?
A- When social media analytics began, the focus was on what the online audience is saying about the brand, product or service. However, as the quantum of data has increased and a lot of integrated campaigns have started happening, the nature of user generated content has also changed. Earlier, when people talked about a brand, they perhaps complained about it. Today, people talk about the campaign, participating in it, the individuals from a company, performance of a public listed company, customer service etc. Hence, the nature of information affects various verticals in an organization- product and campaign specific information affects marketing and communication, complaints will be looked at by customer service department, information about fraud will concern the internal forensic department etc. So, this information is not only used for marketing but it can also be used to gauge the social impact of the business. This is similar to how the captured offline information is put to multiple uses. Online environment has followed suit. Due to an increase in the volume of conversations and improvement in their nature and quality, organizations, brands and agencies have started using online information for multiple purposes.
Q- What does a structured online search allow you to do?
A- Clients manage customer service online and there are two aspects to it. First is monitoring and engagement and the second is work flow management i.e. as the number of online requests increase, the team size increases. Dashboard improves work flow capability by raising tickets which means identifying the conversations tagged around a brand or product and the people in the team who need to address it. The search for a keyword on most dashboards will tell you the article type, the platform on which maximum conversations take place, for instance ‘micro blog’ indicates Twitter is the platform. If I were a brand trying to strategize my social media offering, I’d create a universe of keywords (related to the brand) and leverage this kind of information. In a lot of cases, this may not be happening because we blindly assume that Facebook is the preferred social media platform in India, which is true to some extent. In reality, Facebook tends to do well in a B2C environment but globally, not in a B2B environment barring a couple of companies which have strong content based strategy and hence, pull it off. So, as a brand, you need to develop a strategy for the platform where maximum pertinent conversations happen and analyze the nature of those conversations. If people are talking online about your recently launched campaign, you should engage with them. On the contrary, if they’re slamming the product for its service, then your customer service team must take that online information into account and act upon it. Then comes the news, which unfortunately, a lot of people and PR agencies in India haven’t looked at. You need to manage certain key individuals in media who can influence how your brand is perceived by what they write. Similarly, identify the media houses that are relevant to your industry and trade. Focus on the ones that don’t write much about you online as compared to your competition, especially when they print about you offline. Note that whatever goes online doesn’t go offline due to space constraints. Next are blogs. Blogs tend to be detailed which means that people think through before writing and elaborate on their thoughts, unlike tweets. People may comment randomly or copy information from somewhere but you should analyze the blogs of those who genuinely share their opinion about your brand. If such people are writing good things, then these may be the people you’re looking for. On the contrary, as I call them ‘badvocates’, if they have a detailed negative opinion, then you slowly need to address their concerns to salvage your online reputation.
Q- What should be the choice of tool for search and analysis?
A- Your choice of tool is based on a few parameters. One is the business objective and if a tool can deliver it. It also depends on the bandwidth you have i.e. if you have a bandwidth at the back end where there are people who can manually tag the data, clean it, put a presentation together for the senior management etc., then you should go for a tool that offers more raw data and not too much analysis. On the other hand, if you want a final analysis of data, then you need an advanced tool. The third parameter is budget or cost. There are tools that begin at $100 and some at $500. Radiance6, Sysomos are prominent tools which are at the higher end of the bandwidth and their monthly expenditure is high if you have a large universe of keywords. So, if you compare the cost of Randiance6 and ThoughtBuzz at the basic level, the difference is of around $400.
Q- When should you graduate from using one search and analysis tool to another?
A- It depends on the social media implementation stage you’re at. If you’re at a stage where you’re still exploring whether you should get on to social media, then tweet-tag is a simple way to start with. I think we’ve crossed that stage where a simple tweet-tag could help you. Today, getting on to social media is not a question of if; it’s a question of when and how ready we are. So, you need to understand what the environment is all about. A tweet-tag only gives you a search or monitoring around a particular keyword and doesn’t provide any form of analytics unless you manually measure the number of tweets over a period of time. This can’t be used by any of the bigger brands as manually counting the conversations would be very difficult. Today, any brand would need a basic level of analytics or a tool due to consumer generated conversations across various platforms. For instance, we have been working with a telecom client, whose name I can’t disclose, for 2.5-3 years now. In 2009, when we started working with them, social media was new and so, the mandate was to just monitor the conversations and understand the overall brand sentiment and buzz. At some point in time, they asked us to monitor the competition too. So, for one year we continued to monitor the complaints, feedback, their prices and value added services viz-a-viz the competition. By 2010, people in India had started realizing the significance of Facebook and Twitter. So, the brand decided to use advanced analysis and we came up with a specialized tool to cater to a large volume of data.
Manas– So, what I understand from this is that first, a specialized tool is required when you have a huge volume of data. Further, there are two aspects to it. One is the flux of the volume, for example the number of tweets per hour or per day. The other is the time period for which you need to collect and analyze a volume of data and compare it with what’s been done in the past. Secondly, for effectively managing the reports, analysis etc., you need a specialized tool rather than a generic tool like tweet-tag. Therefore, for big brands, social media analytics has become necessary.
Shamir– To add on, there is another reason to have a specialized tool besides the social search. Conversations are not limited to Facebook or a particular platform. Customers can give their feedback anywhere, may be on a niche blog that you’re not aware about. A tweet tag will only give you tweets whereas a specialized tool will scan the entire web and give you information from anywhere across the web. Today, start-ups monitor conversations online as they can’t take the risk of listening to one platform while their customers may be on another platform. Interestingly, VCs and Angels have started monitoring conversations online to figure out what’s happening in this ecosystem, the budding start-ups, the hot spaces etc. It has really evolved from listening to simple chatter to trying and understanding the insights in a specific domain or sector.
Manas– Being comprehensive is one part of the specialized tool that you cannot get from a simple tweet-tag. Also, it’s not always about searching for your brand. In general, if you want to analyze and make sense of the industry and/or the competition, you will need the specialized tool.
Q- Although you can’t disclose the names of your clients, can you give examples of companies that have used social media analytics?
A- We have examples from automobiles, to financial services, to banking, to F&B etc. I’ll give you a particular example. An automobile client was about to undergo rebranding as part of their marketing mandate. They weren’t sure about how their rebranding will be received by the consumers, industry and others who followed the brand. I look at it as post-campaign measurability or feedback. As part of the integrated campaign, since they were going to do the mainstream market research to understand how people perceive the new brand, they decided to do it online too. It was easy as we just had to monitor the buzz and consumer sentiment. Then came the tricky part. They decided to correlate it to sales and that was a tad different and difficult. It changed the scope of work because so far we only had to measure their brand but to correlate to sales we had to monitor the competition also. In automobiles, there are a lot of categories depending on how the car is positioned. So, we had to look at various categories. Also, we had to understand the sales of the competition and if it correlated with the online buzz. It challenged us but ultimately we found perfect correlation between online buzz and the sales figures, not only for the client but also for the industry. It was an eye-opener for both the client and us as this was the first time we were doing a project like this. So, the client realized that they could increase their sales by spending some more money online. They were spending money on traditional digital outlets like advertising, campaigning, newsletters, mobile space etc., but not much on social media. They did it for three months as trial and after that period the figures were again impressive. Slowly, they got convinced that as they had started to spend slightly more money on social media space, the sales figures had started to improve. Now, in the automobile sector, they’re probably one of the pioneers in spending money on social media. Barring two quarters, we’ve been able to show them every time that the money they spend on online space (ROS) gets translated to sales. To be fair to the client, in certain cases they did standalone social media campaigns. In a lot of cases, they did integrated campaigns where may be the execution of the campaign was online.
Manas– So, social media analytics is also a way to find out the general ROI that you get on your social media initiatives.
Samrat– Social media analytics can work on multiple levels. I can use the user generated data to plan a campaign or can use it post the campaign to understand the type of buzz created and the acceptability of the campaign.
Q- Can you also give an example of a small company that used social media analytics to drive business results?
A- Customer acquisition is a challenge for any startup. In fact the entire business revolves around customers. So, it’s crucial to understand what the customers are talking about, their needs and feedback. Today, people are sharing their experiences online, both good and bad. It’s important for a new start-up to digest the overall customer sentiment. We’ve been working with a few online retail apparel clients, some healthcare clients which are start-ups etc. If you look at the case of healthcare product start-ups, they’ve been monitoring the online platform to understand where the customers actually exist. They had two modules- doctors and patients. They wanted to know where the doctors engaged on social media. Was it Facebook or a particular blog? Who are the patients talking about healthcare issues? What sort of healthcare issues- related to dermatology, radiology etc.? The startup did this exercise for 3-4 weeks to understand their domain better. Then, they channeled their marketing efforts and spent on those niche customs. Now, they don’t have to validate their marketing strategy again and again till they figure out what’s right. So, they’ve saved themselves time by narrowing down the right customer segments and ensured that their marketing is effective. This is one way through which startups are monitoring to acquire customers. If you look at a more evolved stage in business, say a startup that is 2-3 years old, Facebook happens to be one of the more favorite channels especially for Indian companies, where they’re trying to create a community. Brands are trying to identify their influencers, engage with them on a one-to-one basis and incentivize them.
Manas-So, using social media analytics techniques, startups are able to locate where their customers are present. They can identify the channels where their stakeholders are and engage with them. They can analyze the nature of conversations too and tune themselves accordingly. These social media tools help them derive meaningful information from incoming raw data.
Q- How is analytics accommodated in the social media budget and how long does it take for a company to get results from it?
A- Analytics should be allotted 10-20% of the annual social media budget because there are new users coming on board everyday. Social media environment is fast paced and changing unlike the traditional environment that is settled. I suggest such a high figure because even the best of tools require manual intervention and interference after a point. Acquiring data is one thing, obtaining analysis is another but using that data and analysis to take business decisions is the third aspect that most people tend to ignore while planning the budget. So, you need the right people who understand data and business well enough to leverage the former for the latter. Hence, the cost of manpower must also be included in the budget. I strongly believe that manual intervention in analytics is very important.
To answer the second part of the question, it is not a very long process. You have to identify what you need to measure. Once that is done, get an agency on board or get a dashboard. Begin with the universe of keywords that you want to monitor. Depending on the amount of conversations or user generated buzz, you could get the first set of results in 15 days. If the conversations are fairly large, it could take a month. We’re able to provide analysis in a 30 day window in case of standalone projects for clients who just want to know what’s happening around the brand, using data that goes back to 6 months. The only challenge here is that it doesn’t have much of Twitter data as we don’t have firehose for Twitter and it doesn’t allow historic search beyond 4-5 days.
Manas– From what I gather, analytics should be a part of social media strategy because data cannot be managed manually beyond a threshold. Data is like money on the table. If you’re not serious about data, you’re losing money on the table.
Q- For analysis, how to identify what needs to be measured?
A- In general, you should measure the number of people talking positively about your brand, number of people talking negatively about it, the platforms on which conversations take place, impact of a launched campaign in the online environment- if it is peaking and how, the people who are talking about you etc. For specific platforms like Facebook and other sites, you can define a matrix of 50 odd parameters to keep record of people who repeatedly engage on your Facebook page, the kind of media they share, the type of questions they answer- a poll, multiple choice or subjective questions and the kind of media that gets maximum responses. On Twitter, you could measure the number of retweets for a message, for instance Justin Bieber got 50,000 retweets for his ‘thank you’. Further, you can measure how well your brand engages with customers. Whatever is done to measure the impact of the business offline, can be done online too. It’s that integrated and evolved now.
Q- Are all types of measured data pertinent to a business?
A- For any business, it’s important to realize what they seek before doing anything at all. A startup will have its own KPS and matrix to achieve monthly or quarterly results. The number, volume and velocity of tweets are irrelevant unless you know how it will translate to your business objectives. So, you need to know your business objectives and the challenges you face. Then, we guide the clients about the metrics they need to use and how they will serve their purpose.
Q- Can you give some more examples of how measured data becomes actionable?
A- We worked with a beauty product company in the past. They wanted to start engaging with social media to involve key influencers, carry out press releases, product announcements, experience sessions, celebrity interactions etc. They specifically wanted bloggers. The idea was to become more visible online. They didn’t have a social media strategy or a big budget and hence, wanted to start small. This was their business objective. It meant that we had to identify the influencers or individuals talking about their beauty products. After looking at the data, we recommended them to help these people take their online presence to the next level by sponsoring their blogs or providing some kind of technology support. Though it couldn’t be implemented because of the budget, we met their original brief. We identified the individuals for them. Then, they were able to merge online media with their mainstream media planning.