I wanted to write about gamification in the enterprise this week, but big data and digitalization kept showing up in my feeds. Bernard Marr, a very well respected business consultant and a leading voice in the taming of big data, had a tweet series on tips and rules for big data projects. One of his tweets caught my eye, and I responded. Here is the thread:
Meta expands on something considered base or fundamental. A person is fundamental. Size, language spoken, hair color, political affiliation, are all meta to help us better communicate with that person. NSA uses communication links, Jack calls Jill, as the fundamental element. Phone numbers, location, and duration, are all meta of the link. We have seen the power of meta analysis in the papers lately. So what does this have to do with big data?
So, technically then, big data represents all of the meta an enterprise generates during the performance of executing strategy. Ok, but performance is too high level to be the fundamental thing. What is performance made of? I’ll cut to the chase and throw out where we are rapidly going, the digital task. Those digital tasks are expressed mostly in e-mails today, but more and more, they are expressed as discreet, and mostly structured data from online forms. These forms are present in help desk tools, project management tools, workflow software and our own PPM tool, PPM Central. However, for the meta to be useful, it has to describe something more fundamental and not one off. For meta to be effective, it must illustrate differences, trends, and forecasts. Something more fundamental that can be equated to other things based on some meta pivot must be identified. This is where the lexemes come in.
Lexemes are basic units of meaning. Traditionally, they are single words, but in the context of the enterprise, they are basic process steps. Develop requirements. Write code. Call prospect. These basic meanings are in our emails and our form based tasks. And together, they form a lexicon unique to an organization. While many lexemes are shared across business verticals, lexemes specific to donut making are not typically found in an aerospace context. Thus, understanding your lexicon, and its constituent lexemes, seems to me to be the most important digital transformation effort to take on. If you can use meta to fully understand performance at the lexeme level, you have by definition linked strategy to performance. For the first time since our transformation to a services economy, we can model ground truth under the lens of strategy.
So, starting with data is not a bad thing, as long as you are focused on the nexus of the meta that makes up all of your big data. If you do not know from what well all of the meta has sprung, you will get lost. Oh, and by the way, when you focus on lexemes like we do, you can develop algorithms that build new project plans from old similar ones, you can alert people to similar work occurring in the enterprise, and you can even start mining process. PPM Central has always shipped with a full NLP engine that identifies enterprise lexemes. In fact, all of our patents concern lexemes in the project context.