The Role of Machine Learning
Machine learning is pretty much what it sounds like — a computer or machine's ability to "learn" over time, based on information gained during use. For example, when Google ranks websites, it does so using an algorithm incorporating a machine learning ranking signal called RankBrain, which "has been confirmed as the third most important ranking signal for generating search results."
RankBrain's ability to learn on its own is important because, although the algorithm can be programmed to recognize something like page load speed as a factor for ranking, no programmer could account for every possible search scenario. As Searchmetrics reports, "Of the billions of search queries Google deals with daily, it estimates that 15 percent of these queries are new and have never been seen before. RankBrain helps to generalize these unseen or new queries to something it knows how to deal with, enhancing its ability to filter and parse search results."
Google doesn't tell us how, of course, but it's likely that word vectors are part of their secret. What are word vectors?
Word Vectors Drill Deeper
Word vectors use machine learning algorithms to automatically identify "relationships between words and their meanings using context." The idea is that you can train the algorithm to recognize similarities between words and make new discoveries.
For example, NetBase trained Word2Vec to find fan-constructed words similar to "Beliebers" — the affectionate term for Justin Bieber fans. And it came up with "'Selenators' (Selena Gomez), 'directioners' (One Direction) and 'arianators' (Ariana Grande)."
Finding these relationships means brands can find — and target — additional audience segments that they’d be unlikely to discover on their own.
So Google is likely using word vector adjacencies to inform RankBrain results. And one day the connections this technology creates will be baked into social intelligence platforms, much like the way we understand basic demographics online today.
But you’re not a machine, so how can your business discover its own adjacencies right now and categorize, capture and recognize unlikely connections across its own business units and processes? Connections that will ultimately power that next big idea?
Making the Adjacency Adjustment
If you don’t have one already, you’ll want to create (or purchase) a solid knowledge management system. And beyond capturing and categorizing thought leadership, you’ll want to start moving your best ideas forward — so an innovation management platform could help as well. But the best software in the world won’t help if your organization doesn’t experience a cultural shift and make an “adjacency adjustment.” Here’s how:
Lead by example
"When we take a deeper a look at the root cause of [the silo mentality], we find that more often than not silos are the result of a conflicted leadership team," says Brent Gleeson in Forbes. If your leaders don't have an adjacency mindset, how can the rest of your company be expected to? Whichever systems you adopt internally, make certain your C-suite is using them and talking about using them company wide. If you can’t secure executive buy-in, your initiative is doomed to failure.
Hold a Work-Out
Former GE CEO Jack Welch initiated work-outs 25 years ago as a means of creating the "boundaryless organization" he envisioned. This is similar to a Hackathon. The planning process is largely the same for both, as is the end result (assuming you follow through): “wrangling idea management for maximum efficiency” while showing employees their ideas are both appreciated and essential. It’s powerful.
Use Soft Skills
Some people are good with numbers, or spreadsheets, or code — and these are all great skills. But the "ability to influence people outside the bounds over which you have authority" is just as crucial a skill for fostering collaboration across business units. Encourage leaders with this gift to put it to work. Consider having one person dedicated solely to idea management, if you can spare the headcount. Some businesses have entire innovation teams on staff, and experience an exceptional return on investment doing so.
It’s something to explore.
Adopting an adjacency mindset may seem like a low item on your priority list, but it shouldn’t be. Each day you could be missing out on that next big idea, or that new market segment that your competitor just discovered. And keeping pace with where machine learning is headed is never a bad idea. Trust that your employees have valuable information to share — all you have to do is sort out the best way to listen.
This post originally posted on CMSWire
Title image by Linsday Henwood