After chewing on the hare for a while, and believing that most people probably think that my posts on the maturation of the social listening market may have been a bit overcomplicated, I figured it would be worthwhile to follow them up with an example that applies those concepts. The main point of that series was to illustrate that each and every thing that goes on in a market is related to the framework by which it operates. The reason conceptual frameworks are important to me is their ability to help me recognize patterns that ultimately help me influence others as well as understand their needs.
So how could I possibly apply that kooky model to anything but a social media dissertation? Well, let’s start with the model again. Below is an illustration of the “4 pillars” of social listening. The explanation of them is in my previous post. I put them here for reference. The market essentially has developed through people’s desire to first worry about features, then content, and now they care about accuracy. So if we accept that the social listening market is all about accuracy today, let’s talk about one of today’s hottest FEATURES… Geolocation.
In the last 9 months, I have heard more and more people request the need for Geolocation. As people are getting their heads around how to apply social listening, it would make sense to want to know WHERE people are talking about a topic, trend or brand. It helps one to understand where the conversation is concentrated, know if they are talking about your business at a local level and ultimately where you might push content to them at the right time. No one would argue the importance of having this FEATURE (and I capitalize this on purpose). It is going to be a critical need to listen appropriately. But as with the maturation of the market, when a FEATURE becomes the highlight of a decision, mistakes are made. And frankly, this is what is happening on the market right now as it pertains to geolocation. I am currently working with customers who are turning off when they are unsatisfied with the response they get on a company’s ability to deliver the Geolocation FEATURE.
So let’s be clear, as an operational definition, the FEATURE is the means by which a technology can express something. In social listening, geolocation is a FEATURE that allows a person to dig into where the content was created. And from a use case perspective, it can help with things like social selling/purchase intent (I am standing outside a store…hit me with an offer), customer service (I am pissed off sitting on the tarmac), or a marketing campaign (where did I fulfill that promotion) to name a few. So your friendly neighborhood listening company tells you they can give you geolocation down to the DMA or Zipcode. What do you do? You fall in love with the FEATURE, don’t you?
Well I am here to reiterate the concept of the market’s maturation. For other FEATURES and use cases that were critical in the past (share of voice, gender, domain breakdown), I would argue there are many people with very popular social listening tools that have great FEATURES with tons of CONTENT but their ACCURACY sucks. Therefore you can’t take action on your social listening efforts without spending boatloads of time manually coding all that social CONTENT produced by that FEATURE you fell in love with in the first place. You now crave ACCURACY (ever seen the face of a social media manager who has to deal with poor accuracy? They play eye winky games with their boss when they see they might actually not have to stay up all night sorting through junk to get to the answer). In order to make decisions with social data, you need a way to be sure that the data you are using to make those decisions is trustworthy enough to act.
WHY ARE YOU BEING FOOLED BY THAT GEOLOCATION FEATURE?
So let’s talk about the market’s newest darling; Geolocation. So someone can show you can you see data down to a very specific location. I would ask you this question about one’s ability to do that; how much CONTENT is there really at that specific a level on average. Not much, I would argue. Really, if you want to see how many people are talking about McDonald’s at any given moment in your zip code, today there isn’t going to be a ton of CONTENT around that specific topic (trust me, when the topic get specific enough there isn’t as much data as one would believe). Sure it would be great to capture the CONTENT that shows you that someone is complaining about something inside your store. In fact, it would be really valuable. But as the social media fog grows (post link here), isn’t the CONTENT on any given topic going to climb? And what has happened with those other FEATURES as the CONTENT has climbed? Yes, you need ACCURACY to be able to make sense of the data. So you see, each new FEATURE you care about will ultimately begin to flow through this framework. First there will be the FEATURE, then there will be the rise in the CONTENT and when the CONTENT becomes unwieldy you will need to be to ACCURATELY know what is actually happening.
Let’s play that idea out in a scenario…
A quick service restaurant wants to track the performance of their stores so they make sure their social media listening system has the geolocation FEATURE. So they set up a topic to listen down to the zipcode to make sure they can serve their customers better. Their first order of business is to PULL any relevant data where people mention their brand in a particular zipcode (yes…push/pull again). They do it and are tickled that their great feature shows them the CONTENT that resonates. They find 300 soundbites over a one month period talking about their brand at that local level. Great! We got it, this geolocation FEATURE can help us. Sadly, their system lacks accuracy so of the 300 there are, 150 that talk correctly about their stores. No problem they say, we can handle that amount of data so they manually code it.
One day marketing gets wind of the social team’s ability to geolocate down to the zipcode and thinks; awesome we can socially sell at the store level and make sure that social complaints are handled locally in an efficient way. So what do they do? They create a new campaign that PUSHES hashtags out at the store level. Their program is simple, they create signage in every store that uses a unique hashtag for each zipcode (or even store). When a customer walks into the store there is information everywhere stating, “We want to know you are here and what you think…tweet us as #BRANDYOURZIPHERE”. The idea is, we can keep an eye on things by PULLING what people say at this location. Not only that, we can begin to sort what people think from store to store so we can benchmark performance around specific attributes. And you know what, it’s genius, the program blows up and the average usage of these new store level hashtags drives local volume in the zipcodes and triples every other month for six months. So the 300 soundbites becomes 8100 after six months (some as a result of that hastag and some is just organic). How valuable is that geolocation FEATURE now when you have go through 16200 soundbites (let’s assume the 50% accuracy you could be dealing with).
You can’t and you are back to where you started with any other FEATURE of your social listening platform. Without ACCURACY your FEATURE will eventually fail as the CONTENT climbs. It is just another example of the fact that to make real time decisions with social data you will need ACCURACY to do so.
And while this scenario is only hypothetical, I have already argued on the marketing side that as the social media fog rolls in, that you need a laser rifle to market in that mess. Why would it be any different for listening? We already know that your manual processes with inaccurate social data are causing your organization to question your social team’s ability to show value. So as you sit there being wooed by the next FEATURE a social provider shares with you, ask yourself this: haven’t I seen this movie before and shouldn’t I just get up and ask for my money back?