Predictive Analytics: The key to intelligent engineering
IT support has historically been reactive: wait for it to break then fix it, and fix it cheap.
But this simple approach is no longer fit for purpose. Times have changed, the workplace is evolving, and business users are becoming less accepting of a standard approach to IT support.
And there’s a reason for that. Firstly, with flexible working on the rise and people increasingly using their own devices, the lines between home and work technology have become blurred. People want support to be on their terms, at a time and place of their choosing.
But it’s also about reflecting the business landscape of today. Markets move incredibly quickly now, and the threat of disruption is always on the horizon. Companies need to be more proactive and agile than ever in order to remain competitive.
So how do you develop support models that meet these new and more demanding requirements? Simply waiting for something to break is no longer acceptable. You need an in-depth understanding of your customers’ environments. This is where predictive analytics comes into its own.
And when we talk about analytics we’re not simply looking at device information or call data. You need to bring together all relevant data sets– from machine logs and sensor data to customer usage and trading data – to gain a true picture of a customers’ environments.
By using analytics effectively to develop a proactive rather than reactive service, engineering becomes less about fixing problems and keeping down costs and more about having a real impact, both commercially and in terms of creating a brilliant customer experience (and the two are naturally linked, of course).
Here’s how…
Taking back control
Arguably the most powerful benefit of predictive analytics is dramatically increased control over devices going wrong. This may not sound particularly significant, but I can’t stress enough how much of a positive effect it can have on a company’s bottom line.
Unlike in the past when engineers reacted to failures as they happened, we can now use combined data sets to identify issues before devices fail and massively reduce downtime. And we can be much more specific in what those issues are.
A temperature sensor might tell us that a device’s failure is being caused by its location. But by comparing different data sets we may also be able to see that certain locations experience more issues than others at particular times or when certain users are engaged.
So we can pinpoint the real cause of the problem rather than making an educated guess as we may have done in the past.
We can also identify which device failures will have the biggest impact on productivity and revenue and prioritize engineers’ time accordingly.
Think of a water treatment plant, for example:
If a pump goes down it’s going to cost a huge amount of money and manpower to fix it, not to mention a massive loss in productivity until the device is up and running again.
By using predictive analytics, however, we can see which specific parts of that pump are on their way out and send someone out to fix them long before the whole thing breaks. This is much less disruptive to the plant but also has the potential to saves millions in costs over time.
And what about retailers? Even one till going down can have disastrous consequences for revenue, particularly during peak times. But applying predictive analytics to those devices means a fix can be instigated quickly and intuitively with minimal impact on trading.
And again this has a knock-on effect when it comes to the customer experience. A till going down is likely to cause longer queues and disruption, no doubt frustrating any customers in the shop and potentially damaging that retailer’s reputation or brand.
You can apply this logic to almost any business you can think of where costs and revenues are linked to technology staying up and running (which, incidentally, is almost any business you can think of).
Not a silver bullet
All of this is exciting, but pure analytics will only get you so far. It’s part of the overall story but not the whole.
It can provide you with real insight but can’t give you all the answers. You still need to identify what you want to achieve as a business and then focus the analytics on driving those areas.
And once you’ve identified opportunities, you may still need to invest in the change and processes that will help you realize those benefits.
But that’s not to take away from the powerful business impact predictive analytics can have. While it may not be a silver bullet, it does allow organizations to make informed choices when it comes to investing in technology and processes that will support business needs now and in future.
The future is enablement
Analytics may be nothing new, but today we can go faster and deeper with data than ever before, and this area of technology is showing no signs of slowing down.
Those who invest in the power of predictive analytics now will be amongst the first to make sense of all the data sets that are rapidly growing in size and number. This is going to be key to gaining and retaining a competitive edge in the coming years.
The previous IT model of driving down costs is coming to an end. The future is IT as a business enabler, and predictive analytics will play a big part in that transformation.
Enjoyed this post? Read ‘Four reasons ‘engineerless’ field engineering is the future’