If you’ve ever targeted an email marketing campaign to a specific demographic, assigned a category to your blog post, or chosen a hashtag for your social post, you’ve used data classification. Classification is basically the process of chunking up or organizing your data, into different groups or under different labels, so that you can quickly isolate and bring together all of the things that belong to that group, so that you can do something with that group:
- monitor it as part of a metric,
- investigate it and compare it to other groups,
- work with it, like with targeted marketing campaigns, or
- plan with it.
In this post, I’ll discuss the problems that arise from unclassified or improperly classified data, and give some pointers on how to create and apply your own classifications.
You can watch the video version of this post, or keep reading below.
Unorganized Data is Like a Heap of Puzzle Pieces
You can think about the pieces of your data as pieces of a puzzle, each one giving you a clue to a bigger picture, if only you can place it correctly. If that data has not been strategically organized and classified, that’s the same as dumping out all of your puzzle pieces, from all of your puzzles, into one big heap, and then throwing out all of the boxes.
It’s going to take you a lot of time to sort through all of those pieces just to find the pieces that you want to put together.
Are you able to find all of the pieces?
Did you accidentally use a piece or two from the wrong puzzle?
To prevent these types of quagmires, you use classification strategies.
- You keep all of the pieces of a single puzzle in that puzzle’s box.
- You classify by role or function or type of thing.
- When you’re ready to work on the puzzle, you separate pieces into edge pieces and inside pieces, and then you separate pieces according to parts of the puzzle, like making a separate pile of the sky pieces that you come across. In other words, you classify by role or function or type of thing.
- Most importantly, you’re not afraid to undo the things that didn’t help you reach a successful conclusion to your puzzle, like removing a puzzle piece that’s in the wrong spot.
Classification Strategies for Reliable and Usable Data
Much like the puzzle analogy, you can use classification strategies to promote more reliable and usable real world data. As an example, let’s look at classifying and using email contacts.
I think you’d agree that if I use my entire list of contacts as the basis of my business marketing for everything I have to market, it would come off like spam to a lot of the recipients, because they’re not all interested in everything I have to offer.
Instead, just like with puzzles, I use the idea of boxes to classify my master list.
- People that have taken a workshop from me are collected into one box labeled workshops.
- I can further classify these workshop contacts into people that have taken an Excel workshop, or a strategic metrics workshop, or a dashboard workshop, or a process improvement workshop, and so on.
Now, if I have a new Excel tip, I can email that out to just the people that have taken my Excel workshops, knowing that the email has reached people that are actually interested in that thing, and I haven’t spammed anyone.
Of course, unlike puzzle pieces, these people and other data might legitimately fit into more than one box at the same time.
Additionally, while it’s a bit more complicated than the examples I’ve given so far, you can also apply data classification to free text data like customer reviews to help you uncover things like the customer experience as it relates to different products or different services that you offer.
How to create your own data categories?
The first part is to figure out what classifications make sense for you and your business, which is a matter of knowing what labels or groupings will help you make decisions or take actions. Much like the puzzle, you’ll reflect on the roles and functions and types of things you need to work with, and see if working with them as a group, or analyzing them as a group, makes sense for what you want to do.
But don’t go nuts making up labels!
- You’re making as small of a list as possible. Think in terms of capturing the most important characteristics instead of capturing every bit of nuance.
- And standardize that list of options so that you can use them like boxes.
- If you’re already using tags and other classifications, schedule some time to review them. If you come across classifications that don’t help you, just like those puzzle pieces you put in the wrong spot, remove them.
The trick is to apply this to meet your needs.
Classification for the sake of classification? That’s just busy work.
How will you use data classification?
Will you review, and maybe retire, your existing classifications?
Will you start with your email list? Or will you start by reviewing the types of products or services you offer?
Will you make a clean break from the past, and brainstorm from scratch what classifications you need for the decisions you want to make today?
Let me know by leaving a comment below right now!
And don’t forget to share this with someone who needs it. Thank you!