Speaker 1: This is Jackie Kivers, and I'd like to introduce you to my session called Social Listening Tools, The Good, The Bad, and The Ugly. So, I'm not here to tell you horrible things about different social listening platforms, but rather to give you a better understanding and perhaps a new set of selection metrics for when you decide to license your new social listening tool or platform. A little about myself and my company. So, my name is Jackie Kivers. Besides regularly speaking on social listening and social intelligence, I have been working in the field of social insights and digital strategy for the last 15 years. I'm co-founder and CEO of Convosphere, a social listening agency where we specialize in using social data and working with it in the native language in over 45 languages. So, we are a social first agency, and we put social data in context of other data sets where it's relevant, which could be search data, call center data, web metrics. But first and foremost, we are a social first social listening agency, and so are very nerdy and well-versed in understanding the nuances of the different social listening platforms. So, first of all, I think we're all already on the same page, but social listening is the process of using online conversation to answer business questions with actionable insight. So, the important thing to understand about social listening is that it's a process, right? It's not the outcome. The outcome, however, is social intelligence. So, social intelligence is actionable. It is the application of using social listening insights to make data-driven decisions. It's understanding and applying how to do business differently, given the insights that have been generated. So, how do we get there? You know, it's really important to understand that social listening, the process, and ultimately the outcome, is a collaboration between tools that help us gather the right data and humans who generate the insight and do the analysis and make the recommendations. So, tools help us facilitate the process we use, and then we turn those, the process and the insights generated into actionable insights and recommendations. So, the tools help us facilitate that process and are very useful in doing so. Some tools are more useful in different contexts than others. The other thing that's very important to understand about them is that they also allow us to gather the data in a licensed and compliant way. So, we're not scraping data from websites and violating the terms and conditions of different social media platforms or forums, but they really serve as a great tool to kind of gather the data together to be analyzed and start taking some of those first steps in helping us do the analysis. But as much as those tools are needed, the actual development of insights is a very human process. It's not just a platform doing the work. So, despite what you may have heard from different sales teams from different social listening platforms, social listening tools are big business, and they will happily sell you a license for 12 or 24 months. I think a lot of people may come away with the understanding that social listening platforms are, you know, amazing, and they are, but that they take less human intervention than they actually do. So, I think a lot of times they seem like they're a slot machine, right? You feed in the question, you pull the lever and outcome your answers, but unfortunately, it just doesn't work that way. Instead, insight generation from social data or almost any kind of data remains a very human activity. For example, in the last 24 months, we've seen an unprecedented move towards mergers and acquisitions in the different social listening platforms. So, we saw Brandwatch and Crimson Hexagon come together, Meltwater and Sysomos, Ipsos and Synthesio, and this has really reduced the number of kind of enterprise social media listening tools that have been out there and really leading the industry. And I think this consolidation shows that, you know, there's advantages in some of these tools coming together because they have different features or benefits that they can, you know, work together and provide better solutions to their clients. But it also shows that in some cases, you know, that there are quite a lot of tools out there, and not all of them are highly differentiated. In parallel, we've seen that client's frustration has grown as social listening tools often fail to deliver on the promises and of game-changing insights. Clients realize that insight generation from social data of any kind remains, like I said, a very human activity. And it's not just us in the industry who notice this, you know, the ones who are doing the analysis and delivering the insights. This is actually being called out in industry reports, such as the Forrester Wave, and I think we're going to still see more of this to come. The other thing is that when the social listening tools, you know, they present their sales pitch to you, they're comparing on volume, they're comparing on years of historical data. They're trying to give you kind of metrics, which can be compared apples to apples to their competitors so that you can make a decision. But, you know, agencies just like mine actually license quite a few tools because they don't all have the same access, or they don't do the analysis, or the, let's say, parsing of languages, or have the same depth of data and forms in different markets that could really differentiate them. So, we actually license quite a few social listening platforms, and we do that because we've found differences in them. And those differences are really important to our work, especially because we're trying to work across languages and markets. So, while you may feel like a kid in a candy store with all the options available to you, you shouldn't be using their sales sheets to make your final decision because not all are equal for your specific needs. What you really need to do is make your own list. What metrics matter to your business? What do you want? What do you need? Make it personal to your business requirements. So, I've put together a few things, a few metrics that I found that differentiate some of the tools and platforms that might be helpful when considering your list of what's important to you as a business, or as a team that's planning to use some of these platforms and tools. So, the first is, how does it handle Boolean queries? How many operators are there? Not all of the platforms are equal. Some, like Brandwatch, have an additional 14-plus Boolean operators that are specific to Brandwatch. Theirs help kind of hone down on proximity and nearness. Other platforms have other types of queries available. So, it's important to understand, what do you need out of your Boolean operator? How complex does it need to be? How many subqueries can you use, or do you need? Think about the process and the types of business questions that you'll be trying to analyze. The next would be a set of different metrics, including languages, markets, and industries. The first, not all of the language parsers that the different platforms use are the same. We've seen a lot of differences in Asian languages, in particular, Thai, Japanese, Korean. So, if those languages are used frequently by your business or research teams, it's really important when you're testing out the different tools that you're considering that they can handle your queries in all of the languages equally well. The next is markets. Not all the platforms collect the same volume of data or depth of data from the different channels. Local tools, not enterprise global tools, might be a better option if you're specifically interested in more complex or difficult to reach regions. For example, we've found Uscan really helpful in Russia and Ukraine. The last one is industries. So, industries is important because you may be an automotive manufacturer. You may be a fashion or beauty brand. And not all of these same platforms have the same historical data sets in these different industries. So, some have been working for years with beauty and fashion clients. So, they have been pulling in all of the fashion brand details and have a rich historical data set in that industry. So, it's really important to test it out. And if historical is important to you, and if industry depth and, you know, competitors or competitive intelligence is important, it's really important that you try out some sample queries and really kind of probe the depth of their industry data. Next is how do they determine gender and location? So, this is a tricky one. So, it's not always important. Some of this demographic information can be gained from other tools or other research that you have, but if you're relying on your social listening platform to also provide you this information, you need to be aware of how this is both gathered and assigned. So, if it's not explicitly specific, if they haven't spelled it out and given you real insight into how they're doing this, some tools allocate girls' names and boys' names. So, what if your name is Chris? Who knows where you'll end up? Some tools rely on specific geotagging to allocate location. Others do this by language. So, don't be surprised if you see unexpectedly large conversation volumes in Spanish attributed to Spain, which you may find for that specific product or industry out of sync with other EU5 markets. So, it's really important if you're looking or utilizing gender and location that you understand how this is assigned. The next is how good is the built-in analytics, right? The analysis that it's doing, how good is it? How closely does it match your processes? So, language is both cultural and contextual. Social listening tools often struggle with sentiment detection and analysis. Things like tone, irony, humor, and idioms are often difficult to process. How much do you need to count on the tool applying analysis versus how much time and effort and energy does your team have to apply to doing this manually? You really need to think this through and test it out before you make your decisions. The other thing to understand about the built-in analysis, especially when it comes to sentiment analysis, is that true insight comes from digging deeper. You often can't just rely on the automated sentiment algorithms because the same sentiment algorithm is being applied across industries. So, something, say, in the watches category being called a bad boy is actually probably a great product, but in veterinary health, it would be very negative. So, it would be difficult for the sentiment algorithms to understand the difference between veterinary and watches to really apply kind of true sentiment detection there. If you're looking to go deeper, you could use social behavioral models like Plutchek's Wheel of Emotions, which provides a framework for deeper analysis. So, take a concept like cancer, right? Often, I think it would predominantly show as negative or neutral in the sentiment detection algorithms in a lot of the different tools. However, using an understanding of the emotions that people are expressing at different stages of the patient journey, you can really go beyond positive and negative to understand the sentiment, an emotion that sits behind that. Once you understand that, you can truly understand the drivers and the barriers to things like treatment or adherence or switching medications. Sure, everybody's going to be negative about cancer, but a patient could be really optimistic about a new product or hopeful that a new treatment is going to work. So, the more that you can understand that in depth, the better that you can address consumer needs. So, these emotional insights can help you understand the drivers and barriers of your different stakeholders, whether it's understanding purchase intent, brand choice, user adoption, switch in loyalty, adherence or churn, or even customer satisfaction. Going beyond positive and negative can only benefit you. One of the last things to take a look at is how easy is it to get the data out, right? If you're doing all of your analysis in the tool and you're not planning to export it to Excel or work with it in some other platform, great. But a lot of the tools have limits. So, for example, I would say for us, about 25% of a project is done within the tool and about 75% is done, you know, using kind of human analysis across multiple tools. So, whether that's pulling it into another system to look at the demographics or if it's pulling it in to Excel or bringing the Excel into a PowerPoint chart, there is a lot of activity that's happening with that data. So, also, if you want to put social data in context of other datasets, like your own campaign data or call center or sales data, you can do this and it's great because it can give you kind of a more holistic 360 view, but only in as much as you can actually get the data out of these social listening tools. So, some tools have API integrations with BI tools, so you can pull it into Tableau or other tools like that, but some also limit what you can export. You can only get the volume counts or snippets or some limit your export to CSV or Excel files to like 5,000 per day. So, having a true understanding of how you're planning to work with this data, what other tools you need to integrate with this dataset, and how you plan to manipulate and put it in context of other datasets, is really important that you map out that process or potential workflows. So, before you license a tool, you can really understand where your pain points might be and where kind of key differentiators are for you in selecting the tool that you plan to work with. The last one is taking that one step further and using these tools in combination to enrich your dataset. So, traditionally, social listening tools have been backwards looking. They look over the last 12 months, 24 months, historically, but by layering tools, it allows us to enrich the dataset and create new opportunities for exploration. We can create segments and follow online panels longitudinally to measure the impact of campaigns, share a voice, or changes in perception over time. So, doing the social listening, if it integrates with other tools, such as audience, then set up these online panels and listen longitudinally to these audiences over time. So, there's lots of opportunities when thinking about what's important to you as a team or as a company when making the selection, and there's metrics that you'll be provided, but it's really important that you think through who you've got on your team, who are the key stakeholders involved within your organization, who are not just making this purchasing decision, but who will also be using it, both in the day-to-day analysis or in the applying the actionable insights and recommendations, and involve them in your process of mapping out what are the key metrics that are really important or critical for your business to really get the most out of these tools, and to make them the right tools for your team to work with. So, just recapping here, understand your use cases to define what metrics or features are most important for your business. Ensure the right stakeholders are included in the decision-making. Leverage tools that have additional industry efficiencies. Don't expect the tool to deliver insights. People do. Take the data out of the tool and put it in context of other data sets, and don't be afraid to integrate or layer tools. Thank you very much, and if you have any questions, I'm happy to follow up later.