Functions within companies that service end-user needs have historically relied on their customer’s transactional data to determine interests, habits and use of existing products and services. The analysis is is often pretty straight-forward (the customer bought tires 4 years ago, at 12,000 miles per year, they’re ready for a new set), or e.g. collaborative filtering; if you suddenly and consistently start buying products for infants, they compare your profile to others who’ve previously done the same thing, and recommend whatever the other cohorts of your group have purchased.
All of this is fine, and it works reasonably well, but it is all based on a data set driven specifically by the end-user and their actions. Where its going to start to get much more interesting is when the network that surrounds that user begins to provide ancillary data that delivers a much more subtle context to what is going on, and more importantly, what is likely to happen next. Predictive analytics has been around for quite a while, but like most analytic models, it is primarily based on end-user behavior projected forward, rather than factoring in input from the overall environment in which that user is operating.
Now that our environment has evolved to become a transaction enabler, the applied examples of IoT become potentially endless. While IoT is likely to touch everyone everywhere, the deployment of the service depends on the implementation framework, for example:
Location framework
Anyone carrying a mobile device is already geo-location enabled, and the same thing applies to most cars built in the past few years. So how can IoT drive stronger customer experience in this context? BMW is deploying technology that is triggered when your car hits a wet spot on the road. Your car starts to hydroplane, the ABS systems lock in, skid avoided. But then your car sends a signal to all other BMWs in the area saying “this spot on this road is wet, engage ABS automatically as you approach.” A car that is one step ahead of you and focused on your safety is a strong competitive differentiator, and one that delivers a very high value customer experience, even when the customer has no idea its happening.
Temporal framework
Most food has a limited shelf life. The easiest example is milk; two weeks after you’ve bought it, best to not open the carton at all. Right now the only way to tell if milk is spoiled is the sniff test, which I’m pretty sure most people would prefer to avoid. RFID sensors can be embedded in the carton that are set to trigger a signal after the expiration date passes, and goes to a grocery list app that builds out as your shopping date approaches. No more rock/paper/scissors with your spouse to see who sniffs the milk carton. Same premise can be applied to medications as they approach their expiration date, or essentially anything you put in your body.
Network framework
Disaster preparation and response has always been a major challenge. Even with a lengthy heads-up disaster (e.g. hurricanes) people are still caught flat-footed, government agencies struggle to cope, and life sucks a lot more that it should. A big part of the problem is knowing what is where when, and who needs what how quickly. Embedding sensors in all supplies (down to individual items) and having it tied to a visualization app can let relief agencies track exactly where critical supplies are (rather than shipping truckloads of unlabeled boxes, which need to be opened and catalogued on site, usually under sub-optimal conditions).
As mentioned in earlier posts, IoT technology is an incremental capability layer on top of supporting technology like internet and wireless. Because this capability has been around for so long and is so ubiquitous, people are rarely off the grid. Your day to day life is a series of network-enabled transactions; the average person checks their phone over 250 times per day (while for the average teenager, the number of times they check is beyond the scope of current mathematics). Knowing people are creatures of habit, and knowing these creatures are on the grid, provides an environment where everything can be tracked to deliver a far more holistic and engaging experience than we’ve experienced to date. This is particularly the case when things can now talk to each other on our behalf while we obliviously go about our daily grind.
IoT is by far the most pervasive change we’ve seen in the technology space since the introduction of the Internet itself. While this “always on, always watching” scenario can be unnerving for some people, most IoT data is targeted towards the end-users best interest, since the company delivering them is looking to expand their customer’s engagement by adding value to their daily routine. This is good for the provider, good for the end user, and will continue to expand exponentially as more and more devices interconnect.
Summary:
Predictive Analytics Requires Data Beyond Transactional Data
The US is currently experiencing one of the tightest labor markets in history. For well over a year, the number of open jobs each month has been higher than the number of people looking for work. According to the latest figures from the Bureau of Labor Statistics, there are currently 7.3 million unfilled jobs in the US. As such, many companies have started to become more open-minded when considering candidates. Instead of emphasizing the need for specific titles and experience, organizations are now focusing on the skills that a potential new employee may bring to the company. According to LinkedIn data, there are at least 50,000 professional skills in the world. But which ones are the most important? Well, that 50,000 figure comes from a larger report, for which LinkedIn analyzed hundreds of thousands of job postings to determine which skills companies need most in 2019. It was found that today’s employers are looking for both hard skills and soft skills. LinkedIn says that, due to the rise in AI, soft skills, in particular, are becoming increasingly important “as they are precisely the type of skills robots can’t automate.”
