One of my roles as a “management consultant who happens to do tech” is to go into organizations and look for ways in which things can be streamlined and make information more accessible. Organizations of all stripes struggle with this, and periodically it’s good to look at things with fresh eyes.
Mapping the flow of information processes is a sub-discipline all on its own, and over the years organizations have subtly (or not so subtly) changed their flow approaches to match the software they’ve used. Maybe their original flow was actually created by their ERP software, and the long tail of those behaviors continue way into the future, embedded in the organization.
As the democratization of tech continues, we’ve reached a fork in the road about how we search and manipulate Data, with a capital D. We have structured data (mission critical and actionable), and unstructured data (ephemeral but still useful).
Structured data is the purview of database administrators, data managers, and developers when they build apps that use that data. With structured data, for decades the main way to accurately search for information relied on how you coded it going in: key fields determined what you could search for. In the early days when a mainframe computer only had a whopping 1Mb of RAM, key fields were the only way to crunch data efficiently — no way you could load everything in there, much less search it all. If whatever data you were looking for wasn’t assigned as a key field, so sorry.
At a previous organization with a strong marketing focus, we were investigating a partnership with Kantar (formerly Yankelovich Partners), a data analytics company with a great product called MindBase. MindBase is a market segmenting, profiling and targeting platform that contains information on over 240,000,000 consumers in the United States (72% of the total population). Peers at other member organizations were using it as a part of a new SalesForce-based package that enabled them to classify their current and prospective customers according to membership in different marketing cohorts: like Baby Boomers, Millennials, and Gens X and Z, but with finer differentiations. Ask anyone: the ways to market to a Boomer are different than my Gen Z teen. With such a tool, they (and maybe we) would be able to craft very specific messaging to these segments, and devise A/B testing to groups to further sharpen our message.
But alas: our big donor database that has grown for over 20 years did not capture ONE KEY FIELD to make this possible: the personal cell phone number. Even if we started capturing it now, it would take years to build up a suitable number of records to classify our entire target market population. “For want of a nail…the kingdom was lost.”
Unstructured data has long been the domain of so-called Information Workers, which is all of us in offices who look at screens all day. Unstructured data is all that other stuff: documents, lists, emails kept on either personal computers or stored on servers. If you’ve ever had to search for a filename on your Local Disk under Windows, you know it’s a rather blunt instrument. You get successful hits enough to keep using it, but only as a last resort.
But there is HOPE: when you migrate all that unstructured data to cloud platforms, like Microsoft 365, all that stuff is now stored on a platform where such unstructured data all lives together. Emails, threaded conversations (Teams), documents that live in OneDrive, and documents that live in the great file server in the sky: SharePoint. All can be searched (with pretty good results) because part of the intake/upload process of all this stuff is to index it. Build key fields out of everything, because now disk space and RAM are essentially free to us. Cloud providers are buying in bulk and we’re renting that space in little slivers: $12.50 per user per month.
Suddenly, all that stuff that you’ve held on to “just in case” and loaded into your new cloud platform has new life. Those emails from three SysAdmins ago, that you never had time to look through for that missing password? Done. Depending on if you bought an add-on analytics package, import that (unstructured) collection of customer feedback and slice and dice it, to find out what they REALLY think of you. Or, with the audit tools you can see that your Marketing chief seems to be sending voluminous amounts of data to their personal site. Hmmm….
In the Four Pillars of Analytics, the 3rd pillar is Surfacing Information, often referring to pulling data into a meaningful format, like a dashboard, for better decisions. Making cloud platforms a repository of all the stuff we keep, and then being able to pull actionable intelligence out of it, is what makes them valuable tools beyond just email or storage. The tool adapting to us, rather than the other way around, makes it more useful.