I remember a time when social listening was the next big thing. Big brands were investing a lot of money using technology platforms to listen to online conversations, specifically to people mentioning their company. Some were using tools like BuzzMetrics to monitor “brand mentions” and “share of voice” over time. Others were using Radian6 and responding directly to customer issues online.
They used extremely complex Boolean searches to ensure they were getting the most relevant results possible. This approach would yield any mention of the keywords in the query within news articles, blogs, forums and social media-both negative and positive.
This is still important today, especially to get a better understanding of the “market conversation” about a brand or a topic. I’d like to suggest a way to uncover insights about a specific audience though – actual people, not the media, not anonymous folks sharing articles of commenting on a Reddit post, and not the trolls either.
The concept is simple. Build an audience first and then listen and monitor to just their conversations. I call it audience listening and it looks something like this:
Define an audience + build the audience + monitor the audience + mine for insights =
DATA-DRIVEN MARKETING & COMMUNICATONS PROGRAMS
Insights can include – what keywords they use the most; what media publications they are reading/sharing; what channels they prefer to use and why, and their interests and characteristics that make them unique from everyone else (e.g. they prefer using Spotify versus Pandora).
Here are a few examples of B2B/tech and consumer audiences:
I get asked all the time how to build these audiences and there are several ways to do it. I love testing out new platforms and technologies to see how their audience capabilities work. Below are a few that I have used and tested.
This is one of my favorite platforms. You can build audiences a variety of different ways. You can find people who are connected by what they follow, where they live, or how they self-describe; or what they are saying on social using advanced boolean or specific URL sharers. You can also upload audiences from a CSV file. Once the audience is built is where the real power kicks in. Affinio clusters the audience into smaller groups, allowing for further, deep-dive analysis on each audience – brands they follow, celebrities they like, where they shop, who they mention the most, which hashtags they use, what they read and the list goes on. The UI is easy to use for marketers of all skill levels.
Many consider Onalytica to be a tool primarily for influencers. I find it to be much more than that. They have an advanced search functionality that combines laser focused boolean search for both bios and content shared. What makes them unique is that they only yield results for those who have published or shared content within the last 180 days. Their ranking algorithm is second to none and the best in market. Once you identify your audience, you can export them into a CSV file but it doesn’t stop there. You can then upload them into their IRM-Influencer Relationship Management tool for further analysis. The only disadvantage with the IRM is that you have to know what exactly it is that you are looking for. You can search for keywords, topics, brands, etc. and the output is a dashboard showing you various data points like share of conversation, brand mentions, etc.
People Pattern is like a Google Search for mining what the audience is saying and sharing. They have a database of several million Twitter users that will allow you to build audiences based on a bio search for keywords, basic demographics, geography, etc. They also allow for uploading Twitter handles and/or IDs as well. They cluster the audiences into pre-identified segments/personas and will break down the data into various topics like news, media, sports, entertainment and technology. The only issue is that People Pattern sometimes pulls in spam accounts or Twitter handles that haven’t been used in years, making some of the data questionable. Also, they only pull in the last 200 pieces of content shared by each audience member which could be problematic if certain audiences aren’t that active.
Socedo is an interesting tool. I was using it differently than what it was designed for. Their core product is about automation, specifically DM’ng prospects on Twitter for demand generation programs. While they do have a variety of other capabilities, I was primarily intrigued about how they are identifying audiences. It’s keyword based and it pulls in data from bios and content shared. What makes them unique from the others is that they cross-reference Twitter handles with LinkedIn profiles, which also shows up in the results. From what I understand, it’s about a 30% match rate which isn’t too bad. The only downfall is that the don’t allow you to export audiences until after they have engaged with your marketing campaign.
There are several different ways to build audiences using Buzzsumo. Their core product is all about content analysis based on a search query. For example, if I searched for “Artificial Intelligence” the platform would display every article with “Artificial Intelligence” in the headline. It would also show me how many shares each article received on Facebook, Twitter, Reddit, etc. You can “view sharers” of each article and export the Twitter handles. They also have an influencer search based on bio terms and lists. So, if you are targeting “accountants” the results will show you users who have “accountant” in their bio and/or are a part of a Twitter list with the word “accountant” in it. Think of the latter as an audience crowdsourced by the community. There is no further analysis once the audience is identified, unfortunately so exporting is the only option.
I have used DemographicsPro for years and it continues to deliver. It hasn’t changed much over since then but it’s still a platform to consider as a part of your tech stack. You can analyze several different audiences like followers of a brand, hashtag usage and you can also upload a CSV file of Twitter handles. The results allow you to see what topics and affinities (media, specific journalists, sports teams, technologies, blogs, etc.) your audience over indexes on when compared to the general population. The higher the index, the more unique the audience. The only disadvantage is the UI. It’s hard to learn it but once you know it, it’s gold.
I have been using Crimson for about two years; and its one of the best social intelligence tools on the market, hands down. They refer to themselves as “AI-powered Consumer Insights” but most of my usage has been for B2B and technology brands. One hidden gem, which I don’t see them talk much about is whitelists. I use them all the time. Essentially, you can upload Twitter IDs into a monitor (I call it a panel) and perform a deep-dive analysis historically and in real-time on a particular audience. I also use filters to track topical conversations and the results are astounding. One disadvantage which I have asked about for two years is the ability to search bios. Not sure it’s on the roadmap but it should be as some of their competitors already have this option.
Currently I use Onalytica and Buzzsumo to find audiences and then I upload them into Crimson for ongoing analysis and tracking. Another way to build an audience is to first perform a conversation analysis on a topic, export all the conversations into Excel, isolate the Twitter handles, remove duplicates and then re-upload into a different monitor for tracking.
In full transparency, I have never used Spredfast’s Intelligence platform. I have seen a few demos over the last few years but it’s not the same as getting my hands dirty in the technology. I am not sure what it can and can’t do as it relates to finding, building and monitoring audiences in real-time but in looking at their capabilities on their web site it seems fairly sophisticated. As it relates to social media management and customer care, the product is second to none and highly recommend it for enterprise companies.
By no means does list account for every piece of technology available. These are just the platforms that I am close to and/or have used in the past. Please let me know if I missed one or if there is a certain platform I should look into.
Holler at me questions and thanks for reading.