Recently over lunch with an analyst friend of mine, the topic of data came up. (I know, exciting right.) Let me skip to the end and tell you our assertion – Data is useless. Coming from someone whose job it is to consult with companies on the tremendous value of data, it may seem a bit out of place, but the message I sought to convey, and that I help companies understand, is that data without segmentation is useless, and as I said over lunch, “It’s Just a dataset.”
The role of social data in enterprise environments is growing much more complex, and part of that complexity is driven from the need for this data to inform other key parts of the business with real, tangible benefits that can guide decisions, predict outcomes and validate expenses. One example of a specific segmentation that I build and deploy for clients on a regular basis is called category segmentations. Allow me to provide an example. If you’re a market researcher, for a major computer company whose job it is to gather consumer insights data, are metrics like share of voice or volume trend going to matter to you? Though they are valuable in some applications, you’re much more likely to care about the “why”, instead of the “what”. “Why do people love or hate the new product my company launched in market?” “Why are we having trouble resonating with a specific target market?” “Why is the product being returned more than projected?” These are the types of questions that are being asked and in order to provide an adequate answer you need to be able to perform segmentations on the data.
The recent launch of category segmentation within the Visible platform provides end-users an immense amount of flexibility in modeling the data. If, for instance, a computer company wanted to understand the purchase cycle for a specific product or set of products, they could create a general search on those products and couple that search with a category segmentation that is constructed in a way to pinpoint how customers speak throughout the lifecycle of the purchase. Then, these two searches can be overlaid on one another, and the end-user can begin to see what drives pre-purchase or sometimes called “intent to purchase.” Additionally, segmentations can be built and deployed to capture the voice of the customer at the time of the purchase, and during the post-purchase period, which may be used to determine the highlights and detractors of the product and even what issues that may be contributing to returns or even a potential recall.
Here’s a quick example:
Within VI, you can take the core search listed below, and then you can select any one of the listed Purchase Cycle Segmentation searches to overlay against the core search, and begin understanding how the purchase cycle is represented within social media.
Core Search: “Lenovo”
Purchase Cycle Segmentations:
Pre-Purchase/Intent to Buy: “I am thinking about buying” OR “I might buy” OR “I would like to get” OR “I would like to buy” OR “I am thinking about purchasing”
Purchase Indication: “I bought” OR “I just bought” OR “I got” OR “I just got” OR “I purchased” OR “I just purchased”
Post-Purchase: (“I bought” OR “I just bought” OR “I got” OR “I just got” OR “I purchased” OR “I just purchased”) (“I love” OR “I hate” OR “I like” OR “I really like” OR “I don’t like” OR “I really don’t like” OR “I’m happy with” OR “I’m dissatisfied” OR “doesn’t work” OR “I’m, returning”
If you wanted to isolate each segmentation as a stand-alone, you could overlay them one at a time, or if you wanted one dataset that housed all purchase cycle data, you could simply join all the searches together in one large search to overlay against the core search.
Steve is a Solutions Engineer at Visible, where he is responsible for helping our clients build social intelligence solutions that inform their strategic business decisions and contribute to the successful adoption and scale across the enterprise. Steve resides in Maine with his wife and daughter, and when not at work, has been known to play a little guitar. Follow Steve on Twitter @Reeveseey
