The concept that we refer to as social media listening (also referred to as social media monitoring), is more than just watching for mentions of a company’s brand, products, or services or even looking for customer satisfaction data. Social listening platforms offer a variety of analytics on the back end; if used to their fullest capabilities, can help companies gain insight into consumer behaviors and align marketing strategies to be truly effective. Analytics can provide information to help with:
- Customer segmentation analysis
- Identifying potential consumer groups
- Enhancing Marketing Messaging & Strategy
Depending on the platform, companies can benefit from a range of analytical reports. All listening platforms include basic information about the content posted, while others offer more robust options. The charts and graphs are impressive, but looking beyond this and digging into the data is where the real meat is.
How can social listening analytics be used to capacity? First, it’s helpful to understand how the “extra” information is collected and how it is tabulated. From there, it is easy to understand what reporting options are available and how to drill down in each data point to gain the most insight and actionable data.
Where do the analytical data points come from?
People on social leave a lot of breadcrumbs on the trail – have you ever thought about how much information you provide about yourself when starting a new social site or engaging in an online forum?
Thanks to the social media site design model, consumers are providing a wealth of information about who they are – what do they like? Are they married? Parents? Like to travel? Most social sites require users to create a profile and these profiles capture demographic information on the individual – birthdate, gender, etc. – in addition to the aforementioned interests and activities. As public conversations are collected, themes are built around these personas to further create a customer profile.
For other sites, such as forums or sites where readers post comments to blogs and news articles, the information collected isn’t as robust. Most request nothing more than age and gender. If a phone number is required for recovery purposes, the area code also provides a general potential location for that individual.
Beyond that, the advent of “check ins” or tagging on sites such as Facebook, Instagram, and FourSquare compile consumer data about where people visit and other location-based information.
There are, of course, some caveats. Social platforms can only collect public-facing content. So, for example, if you’re a Facebook user and your profile settings are set to private, then your information cannot be collected in any social listening platform.
Another barrier comes with sites where account setup questions are very limited. An example is Reddit – if a user chooses to sign up via email, the only information captured is username, email address, and password creation. There’s not a lot to go on there, and that’s just how the social site works.
Some forums are open to the public to register and use without much ado. Others are password entry only, which are not considered public facing. In this instance, the content will not be captured, nor will the profile information.
Long story short – not all published and collected content will have user data attached. Does that make a difference? Not really. The good news is that social research isn’t scientific or statistically significant, and it shouldn’t be thought of in that way. Because there are so many conversations that can be captured, the analytics that are available provide enough of a representation to yield the insights and information marketers can use.
It’s fascinating to know that all of these various data points are captured every time you post a funny meme or complain about the weather on social, but it happens. To know that it’s happening is one thing; to understand how these data points and published content are aggregated and knowing how to make the most of the content is another.
Basic Analysis Points – What Are They and How Can They Help?
Social listening programs offer a variety of basic and unique reporting features. Let’s start with the three basic analytical data points – what are they and how can they be utilized?
Share of voice
At its most basic, share of voice is just that – how many times is a brand, product, or service is mentioned on a daily, weekly, or monthly basis. Peaks and pits in volume can be telling – peaks may show the excitement over a new product launch, or on the flip side, a potentially viral situation that needs attention right away.
Pits are also important to note – why aren’t consumers talking about your brand? Are you not providing content, promotions, or information that consumers want to read, share and engage with? Are you not tending to your social sites? Publishing regular content is important – if you don’t build it, consumers can’t engage with it.
Share of voice can also show how many conversations are happening on each type of social site (Twitter, Facebook, Instagram, blogs, etc).
Share of voice by social site is also important to look at. Social media marketing is tricky, and what once used to be a “brands must be on all social sites all the time to be effective” is now more selective. Brands may use the three top social sites (Facebook, Twitter, and Instagram) but really focus their efforts on Facebook, for example, as they believe that strategy reaches the most consumers with the strongest ROI.
Are you sure this is right? Share of voice by social site may tell you otherwise. Continually monitoring where consumers are talking about your brand will help shape your marketing strategy. It may be disheartening to put all of the marketing efforts into Facebook only to find that the majority of customers are hanging out on Twitter. But, knowing this will give you the insight and ability to shift your focus or work to get those Twitter users on your Facebook site, though this is a MUCH harder task. Consumers have their social preferences, and it’s easier to bring the message to them than trying to get them to change social loyalty.
Age & Gender
The information you find here may or may not surprise you. You have an idea of who your demographic is. Does that align with the social conversations surrounding your brand? If so, you know you’re on the right track. If not, does this mean you’re barking up the wrong tree?
Not necessarily. It could signal a misdiagnosis of your consumer demographics, or it could identify a potential new demographic that you weren’t aware of. Isolating age or gender demographics that you don’t specifically target can uncover some interesting information and help learn more about a potentially emerging customer base.
This is perhaps the trickiest data point out there. If an agency or monitoring program tells you sentiment is calculated with extremely high accuracy or there is no manual labor involved, you may want to rethink using that platform. Yes, natural language processing has come a long way in gauging the sentiment of social conversations, but it isn’t (and never will be) perfect. It takes time to review and manually update sentiment settings as conversations are read, and this can be difficult to manage for large brands or topics where there is an excessive amount of content. However, taking the time to work through the results and manage sentiment can be very useful.
Sentiment is generally broken out into positive, neutral, and negative. In many cases, the majority of the content collected will be neutral. This could be someone sharing information about a new location your company is opening or a simple check-in to one of your restaurants. The positive and negative is where the focus should be – what delights your customers, and where are their pain points? Couple this with themes to identify specific areas that need to be addressed.
Themes are just that – words or phrases that are used often when people are talking about a particular topic. Word clouds can be created, showing which words or phrases are being used most often and which words are used in conjunction with other words. A quick glance of the conversation themes can signal a strong like/dislike/focus on a particular aspect of your business, maybe a specific product or promotion, and can also point to potential pain points or issues. Themes can also be compiled by hashtag or emoji – this is extremely useful, as you may be able to track the effectiveness of hashtag marketing or even discover consumer generated hashtags that are being used that you may want to incorporate into your own marketing strategies.
Tips to Make the Most of Your Analysis
The basic data points are great for learning some high-level information about consumers, gauge satisfaction, and monitor potential issues before they become too big to manage. The data can also be used to help with marketing efforts. Below are four examples of how to use the data to its maximum potential.
Go beyond your brand. Listening to online conversations specific to your company is important of course, but think bigger.
If there is a trend that seems to be emerging, whether it’s a new demographic talking about your brand or product trends, it’s time to expand listening horizons.
Social research is an effective methodology to look at what is happening within your industry. It’s essentially the same as a brand’s social listening program but broadening the scope or focus. Take that data and conduct social media research to focus on the themes or products in general (outside of just your brand) and see what consumers are saying. It can help companies have a better insight into consumer perceptions, which allows for targeted marketing messaging and strategy.
Look at sentiment by age and gender.
Overall sentiment is good to understand, but digging deeper can uncover issues and pain points that may go unnoticed otherwise. For example, a brand may be looking at overall sentiment for a time period and see that a good percentage is positive, and quite a bit is neutral, but there is enough negative sentiment to cause concern. Is this general, or specific to one customer segment?
One example is a financial institution that used social listening. They were concerned by an increase of negative sentiment over the last few months. Looking at gender differences, the sentiment pattern was similar to the overall results. However, something interesting happened when looking at age demographics. The institution found that the negative sentiment stemmed from those in the 18-24 age group and was able to isolate conversations in this group to learn more. What they found was that their banking model was one that this age group did not find useful and the specific pain points that were not evident in the other age groups were uncovered. The financial institution was able to learn what this age group needed and create a more positive banking experience for them.
Review sentiment by themes for further insight.
Identifying sentiment as a function of themes or keywords will potentially identify pain points or aspects of the customer experience to promote in marketing messaging. It may be something as simple as a product or service “wish list” item or something more significant as an unknown breakdown when consumers are trying to get issues resolved through the phone system. Picking up on this information feedback can enhance operational standards and improve the overall experience.
Age & gender can identify new customers.
Did the social media listening platform show an age group or gender data point that was unexpected? This may signal an opportunity to expand the consumer base. There are two things to do when this is uncovered: look at the specifics of the conversations in the unexpected age or gender group and do some broader social research to investigate this demographic trend across your industry.
First, single out the conversations collected for the demographic to find out more – what are they talking about specifically? Is it a particular product or service you offer or is it more of a generalized interest? Or is it a simple uptick in conversations about the business or industry?
Once this is determined, set up a filter in your current program to continue to monitor this demographic over time. This will provide further insight and help determine if there is a need for your services here.
Another option, either in conjunction with this targeted listening or after some time has passed, is to expand social listening to collect social conversations around the industry or specific products or services to find out if this demographic is a potential new customer base that you weren’t aware of or if this was a one-off fluctuation.
Themes help brands talk in their customers’ language.
How often have you heard the phrase, “I’m going on a Target run”? Often, right? Do you also hear, “I’m going on a Walmart run” or “I’m going on a Publix run?” Not so much. For whatever reason, Target customers are often heard using the “Target Run” phrase. Earlier this year, Target took advantage of that and launched their TargetRun and Done campaign – with sales in a bit of a slump, the retailer used a popular consumer phrase to regenerate its marketing efforts.
This is something that can easily be done by identifying words, phrases, or even hashtags that seem to come up week after week when listening to social conversations. Looking for consistent phrases, hashtags, or consumer-driven nicknames for your company, products, or services, can give you fresh marketing strategies and messaging.
Social media listening is a basic component of any social marketing strategy. At its most basic, it’s great for monitoring conversations. The reporting features that accompany listening programs, however, can provide richer insight into your consumers, their preferences, and how they view your brand. Digging deeper into the nuts & bolts behind the analytics may give you better insight into your consumers and provide new marketing messaging and strategy ideas.