Six Reasons to Drop Your Social Listening Service
For most small and mid-sized brands, social media monitoring services provide overwhelming data but underwhelming analysis.
Usually New Year's resolutions represent incremental changes—work out more, go out less, finally put that Rosetta Stone subscription to use. But why not consider a more substantial change for 2019 that will save your business money and free up marketing resources? Drop your social listening service.
Sure, these tools have their uses, especially for big brands and e-commerce leaders. But for most small and mid-sized brands—think bigger than mom-and-pop burger shop but smaller than a multinational conglomerate—they're pricey investments that rarely deliver actionable insight.
Let's look at six key areas where social listening services fall short.
Facebook remains the largest social media platform and the weakest link for monitoring tools. This is because Facebook is a "walled garden" that gates off third-party access to its data, meaning that any "monitoring" is limited to posts that have been either made public or come from the friends list of the account you're using.
Given these limitations, the most relevant search results come from posts to public groups, especially advocacy groups, and any conversations and shares these posts generate. Brand mentions in groups (or in any public post) are worth tracking, but they still represent a fraction of total Facebook content.
Facebook search-result volume is so limited, in fact, that's it's not difficult to keep up-to-date on brand or keyword mentions by simply doing the searches yourself. It's worth asking why you're paying a tool to do something that's easily accomplished with a simple date-filtered search.
Instagram is also a walled garden, and thus also a challenge for monitoring. Few monitoring services (if any) offer anything more sophisticated than aggregating posts by hashtag or location. These options aren't going to help you surface the most valuable Instagram data— namely public "Stories"—as this requires knowing the name of the most popular geotag you're searching for and manually watching the Stories every day since these posts obviously disappear.
Although you can search for feed posts via hashtags and geolocation, you cannot search for keywords. This is a huge limitation that leaves out a significant, if unknown, amount of data. Additionally, as a largely inspirational platform, Stories offer a more conversational place for users to share their thoughts. Not having access to this information leaves monitoring services without the tools to do their job effectively.
Twitter's open API is extremely developer-friendly, and as a result, many monitoring tools feature impressive Twitter analytical capabilities. And given the platform's scale and constant stream of data, automated monitoring can be useful for immediately surfacing mentions that could otherwise be missed.
But for the brands most reliant upon Twitter—especially those using the platform as an extension of their customer service—comprehensive management tools like Sprout Social represent the best investment, as they allow service teams to access customer data and review previous chat history.
If Twitter is a critical platform for your brand, and one that moves too quickly for you or your team to adequately oversee, listening tools may make sense. If not, then ask yourself how much value you're actually getting from automated analysis.
The common critique of sentiment analysis is that it fails to pick up on important nuance and context, and thus incorrectly flags sarcastic comments as positive ("Website is down again. Great work guys!") or humorous comments as negative ("More beach photos? Seems like a terrible job!!! 😜").
These shortcomings are widely acknowledged even within the industry, and in fairness, the accuracy of sentiment analysis has improved substantially in recent years. The more important critique, though, is that sentiment analysis isn't telling you anything you don't know or couldn't guess. If your server isn't working or your orders are shipping late, customers are going to complain on Twitter. When you post pictures from charitable endeavors involving puppies, people are going to like and share your Facebook posts.
There is, of course, something to be said about the ability to quickly glimpse and understand the scale of a certain sentiment. But it's worth considering whether or not the volume of mentions is sufficient enough that you really need an expensive tool to cut through the noise.
It's hard to gauge how influential an "influencer" really is in 2018 given the ease with which new followers can be bought and engagement numbers inflated. Even if you know that an account has a large number of real followers in your target demographic, it’s nearly impossible to tell the ROI on #branded content.
Influencer analysis provides the most value for e-commerce purposes, especially if you're engaging in influencer marketing where you're paying social media stars to help you drive brand awareness or directly promote product purchases, and then tracking effectiveness through UTMs and discount codes.
But if your brand isn't involved in e-commerce and you're not engaging in influencer marketing, then you're probably not gaining much value from influencer analysis. For most brands, an influencer is effectively just a person with a large number of social media followers within your industry or target market. You likely don't need a fancy monitoring tool to tell you that information.
Some monitoring tools move beyond social to cover the entire internet or at least a sizeable portion of it. These tools will tag brand mentions or keywords by platform—usually some combination of web, news, social, forum, and blog.
On the surface, this kind of sorting seems useful, but the methodology is pretty fuzzy once you start digging in. There seems to be a considerable overlap between "news" and "web" sites. And what constitutes a "blog" in 2018? Is it any site hosted on WordPress? And what about a forum? Reddit, sure, but one tool we used tagged anything posted on a Medium site as a forum mention.
Problems associated with sorting are less important than other monitoring issues raised above, but they do underscore the fact that so many "insights" presented to you by monitoring tools are based on questionable methodology. Does it really mean anything if a monitoring tool tells you that seven percent of your brand mentions are on blogs or forums if these terms are never precisely defined?
For most mid-level brands, the most important insights from social media monitoring come via active day-to-day listening. That means reading every social post that mentions your company or a relevant keyword or at least striving to do so whenever possible.
This may be impossible if you're, say, Coca-Cola, but many businesses that are paying for enterprise monitoring tools likely have a sizeable but essentially manageable number of daily brand mentions and relevant keywords. If you're making an effort to stay on top of these—and even track them in a simple spreadsheet—what value is a monitoring tool really adding? In most cases, sentiment analysis, influencer analysis, and word clouds are simply bells and whistles that add little to understanding consumer conversations.
Before you invest in an expensive monitoring tool, consider the specifics of what the tool is delivering and whether or not that level of insight or analysis couldn't be provided by an entry-level employee who has real-time access to your business needs and goals or an agency partner with a comprehensive understanding of these platforms.