Social media has grown exponentially in the last five years to become an integral part of the marketing arsenal. Increasing levels of resources are being dedicated to social media, with B2B companies allocating, on average, 10% of marketing budget to the running of these profiles according to Circle Research. Although businesses are utilising these social tools, often on a daily basis, there is less understanding on how to make use of the data received from these channels and more importantly whether the efforts made across these channels are driving the desired results. Social media analytics can help to make sense of this influx of data and ultimately driving not just marketing but all business decision-making, determining ROI and creating the pathway for an effective digital presence.
The Power of Word of Mouth
Word of mouth is incredibly influential and has the power to radically impact the public’s perception of a brand. Social media websites help facilitate this, enabling people to air their grievances or praise in real-time and in a very public way. This highly public nature, and the inability to control the narrative, means these channels can be highly volatile for businesses. With this in mind, it is not surprising that more and more marketing departments have social media strategies employed as part of the overall marketing activity, with 4 in 5 B2B marketing departments now active on at least one platform (Circle Research). Monitoring online conversations relevant to a brand or business enables companies to react instantly, understand what is being said and where it comes from – making it an essential part of online reputational management. Learn about the use of personality in Social media.
‘Social media listening’ is the process of tracking online messages that have been created and curated by individuals over social media networks. This often involves monitoring channels for specific words, phrases, trending topics or brand names that are relevant to your business. This process extends not only to social networks but can also include longer-form content such as blogs and multi-media communications such as vlogs and slide shares. With 2.078 billion active social media accounts in 2015 (We Are Social), there is clearly a wealth of data that can provide companies with intelligence to make effective strategic choices. In fact, social media is becoming increasingly recognised by companies as a powerful research tool, with an industry report indicating that 78% of companies are planning to integrate social media data into marketing campaigns in the near future (Social Media Today).
The stages involved in social media listening, as well as extracting and making use of the data received are discussed in this article.
Listen with a Purpose
Before beginning this process, it is important to consider the purpose behind listening on social media. Going forward without any goals in mind can make it difficult to understand the significance of the data and how it can be an asset to the business, overall rendering the data redundant.
In addition, there should be a level of awareness relating to how social media data can tie into not just marketing goals, but larger business objectives. An example of this could be that an e-commerce company’s primary goal might be to increase sales, in this instance, social media listening can indicate to marketers which social campaigns are creating the most buzz, this can help inform which campaigns receive greater levels of investment.
Gathering Data
A challenge that presents itself during the gathering data stage is the volume of data that is available and the multitude of social platforms you are able to monitor. This means that manually gathering data is an arduous process that is often time consuming and susceptible to human error. Larger companies are often met with a high quantity of data, which can seem infinite and unmanageable, making it a near-impossible task for an employee to execute. While for SMB’s, where resources can be strained, it also presents a cost-to-value problem.
Introducing software to automate this can significantly reduce the time associated with gathering data. This proves to be particularly advantageous as data should be collected at a fast enough pace for it to remain relevant and valuable as changes can happen before conclusions are drawn. Implementing software is also a solution for aggregating listening across multiple platforms in a cost-efficient way. Social media monitoring software incurs a cost for companies but it usually proves itself to be worth the investment, allowing for the allocation of time to be spent elsewhere. Examples of high performing software you might want to check out include Social Studio, Brandwatch, Mention, and Cision.
There are also free tools that can help, on a basic level, with keeping apace with online messages and monitoring conversations. Integrated KPIs (Key Performance Indicators) on social media networks often offer the user the capability to monitor specific conversations, TweetDeck for example, enables users to save a search of specific keywords into a column so that any certain discussions can be seen in real-time. While Facebook Insights reveals the engagement and reach from company posts, aggregating results from previous posts for comparison. There is often the opportunity to export this data for further analysis, enabling it to be recorded and stored by the company.
Organising Data
The next step is organising the data into relevant categories. A big struggle at this stage is quantifying this data from these networks. Communications from all these platforms are cluttered coming in different formats, ranging from posts, tweets, comments and more.
Ultimately though interactions are often categorised through sentiment analysis, or deducing the emotion behind the communications. These messages are ranked as positive, neutral or negative to give companies an overall consensus of public opinion. This can be expanded to specific product or brands to understand why one campaign is working, while another is floundering.
The software has the intelligence to make assumptions based on semantics to determine which of these categories these messages fall into. However, accuracy for these tools tends to around 70-80% (Digital-MR). This is because the way people communicate over these platforms is often not in complete English, with abbreviations and colloquialisms commonly adopted that may struggle to be picked up by a computer. Tone can also play a part in messages being interpreted incorrectly, with sarcasm, for instance, being easy to mistake. However, as these tools are maturing, the latest systems are even incorporating artificial intelligence to categorise data to learn through use so the accuracy of the process increases as time goes on. While unable be completely accurate, precise results can be driven with this software at this stage as it follows the same set of rules each time, acting as a benchmark for future research.
Interpreting Data
Many interactions cannot be isolated for interpretation as it often comes from a much larger data trail or discussion in which the wider context needs to be understood. One negative comment may have followed a discussion of positive comments, for example. Additionally, social media users also exert varying levels of influence to their audiences, which can affect exposure, for instance, celebrities or bloggers, the ‘who’ needs to be evaluated so that greater weight can be assigned to these posts.
At this stage, data-mining is made easier by the integration of other social media tools. These tools intelligently measure these variables, pulling all the available data to highlight areas of disparity or areas of interest, as well as identify and predict future trends.
Action the Data
Now that the data has been gathered and interpreted, it needs to be measured against the initial company goals, with consideration paid to how these results tie to other departments within the business. It is from this interpretation of the data that effectiveness of campaigns can be measured, with recommendations driving decisions related to increasing or decreasing spend.
Overall, social media analysis and brand perception is a constantly moving process that is subject to regular change, as is the dynamic nature of these social channels so the process of collecting and analysing can not be a one-time thing. You will need to perform this process on a frequent basis but if you do, it can provide the foundation for intelligent, data-backed business decisions which have the potential to save you money and help increase sales.