Identify your potential customers by their conversations, not demographics, because personalities cut across social strata, they endure over time, and they are identifiable from social data.
We use social data insights as part of a top-down process which starts from the general and moves down into details. This is a new way of thinking about marketing research, which uses predictive analysis based on customer conversation. It has specific advantages over traditional research methods and allows us to stitch together profiles of individual’s across their social channels.
Difficulties with Demographic Data
Demographics are a labeling method, but the information they contain is insufficient. When was the last time you met someone who introduced themselves by demographic labels.
“I am a 30 year-old male who makes $50,000 a year, works as an account manager in a marketing agency, reads Forbes and the NY times, and is married with two children.”
It doesn’t really flow off the tongue does it? The truth is, these labels are artificial ways we have for defining people of interest based on visible traits. They don’t hook into anything visceral about our lives, those gut decisions, where we make the hard choices which dominate our purchasing decisions.
There are some things for which demographics cannot account for. Let’s take a look at where, and why, they fail to capture the nuances of our customer’s lives.
1. Demographic Data is Horizontal
This means that they don’t capture the intersection of identities our customer’s have and they give us tunnel vision. That is why we end up with such a gap between perceptions Millennials have about themselves and how they are viewed by HR professionals.
This graphic was recently shared by my colleague Amy Tobin in our Millennial Think Tank group. It is clear that there are exaggerations on both sides as the additional thought bubbles make clear.
2. Demographic Data Only Moves in One Direction
Demographics are tied to temporal events; many of these categories are recognition of milestone events in our lives. This is most likely why things like age and job title were chosen. They show an expected progression, but to make that information useful it would be better to do a longitudinal study.
This is because a demographic label is just a Polaroid snapshot. It is a moment captured along a timeline which is going forward. This means that some information is lost to us.
Let’s take our example of that account manager from the introduction. Time has marched on and he has lost his job, and his status changes from married to divorced. He is now part of a new demographic snapshot as a 30-ish male who is unemployed and divorced.
What has our market research lost? The impact the transition to unemployment has had on the attitudes and beliefs of our former account manager.
3. Demographic Data is Not Anonymous
Companies interested in using social data insights should be wary of asking for the types of data which appear in demographics. This makes it easier to assimilate into offline research, but it doesn’t fit with the way people disclose information on the internet.
The only information for which we can determine demographics is that which is self-disclosed through conversation or from a profile. Even then, there is some likelihood that the information provided is false in order to protect anonymity. This is why our research uses algorithms; the only tool in our arsenal able to stitch together a full profile using predictive analysis.
Targeting Conversations Allows Us to Fill in Gaps
We are always upfront with clients about the limitations of our methods. They are not meant in any way as a replacement to traditional research. They are a complement, one which allows us to understand customers more deeply as they present themselves in conversations made available publicly through self-disclosure.
Yet, this data, only 100 Tweets or so, allows us to stitch together a vibrant picture, not of demographic labels, but of motivations. I recognize the important role personality, interests, ambitions, desires, emotions, and other elements of the psyche play in purchase intent.
This is crucial, as communications from companies, particularly in social media, can come across as tone-deaf. Social data gives us the empathy required for interactions which speak the same language as customers.
Social data insights form the foundation of a content or community strategy.
Advantages of Customer Conversation Data Over Demographics
1. Allows us to focus on Individuals and not Stereotypes
When we focus on Millennial attitudes, they are found to be similar to the generations before them. We see this played out in our Think Tanks all the time, but here is one particular incident which has stuck in my mind.
Ryan Pannell, Gen X, shared his opinion of Millennial employees he had encountered who he found to lack understanding of proper business etiquette. Tiffany Daniels, a Millennial, shot back
“maturity is NOT limited to a generation.”
But my favorite quote of the night goes to Mary Tennison:
“Idiots are NOT generational.”
These incidences have risen in volume NOT because of demographic proportions, but because there is more media coverage paid to it.
2. Conversation Data is Not Fixed in Time
Do you know there are algorithms which can analyze text and determine authorship? This type of investigation led to the discovery of J.K Rowling’s pseudonym, the mystery writer Robert Galbraith. The way we use language identifies us as individuals and is more persistent over time than demographic labels.
A quality unique to social data insights is our ability to travel back in time. Content which was written six years ago is as discoverable as something which was written yesterday. Algorithms find that data and, with high accuracy, are able to stitch together the gaps missing when we find information linked to an identity.
3. Conversation Data Doesn’t Require Disclosure of Identity
We don’t need to know who someone is to find insights. This is because only language is analyzed. The extracted information tells us a lot about the backgrounds of people having these conversations, but not necessarily identifying information.
Think of this like a radiologist who looks at an x-ray. They need not have contact with a patient to come up with the proper diagnoses. In a similar fashion, the data is enough, without the identifying labels, for us to process and make sense of it.
Use Social Data Insights to Make Marketing Communications Better
I will always feel uncomfortable using these methods due to the privacy concerns involved. I know this about myself and also about the people I work with here at ArCompany. That is why we put deep thought into everything we do. The issues involved are complex and clients should understand there are boundaries in regards to how this information is used.
ArCompany uses social data insights because the information we receive puts customer needs firmly at the center of our work. We have all seen the power relationships bring to the table and how they ease tensions between companies and customers. This turns what are mundane transactions into brand building events. Our goal, help companies prosper by working in collaboration with their customers.
ArCompany uses social data insights as the foundation of our content and community strategies. Discover more about your customers by contacting us today.
Photo credit: Adult Change Tables Inside sign, toilet, Phoneix Sky Harbor Airport, Arizona, USA via photopin (license).
Susan Silver is a community focused strategist who uses social data insights as the foundation of her work with ARCOMPANY. Her philosophy “Humanity in Data” is informed by a background in cognitive-behavioral psychology. She is making positive change in people’s lives, and the world, with thoughtful communication on behalf of her clients.
So very true! Demographics, while important as part of the mix, only provide part of the story. Conversation analytics actually tell you more about what an individual is thinking about right now so when combined with demographics help with more relevant individual (or cluster) targeting and understanding.
Thanks Steve! I like your point about clusters. I think that is more advanced than most people consider, but (especially in digital) it seems that you need to understand who is talking to who. I have been putting a lot of thought into how we think about these clusters and what to do with that information when we have it.