Cognito iQ has been helping companies capture data from the field through our mobile applications for over 20 years. Today, we work with a wide range of field service customers in industries such as facilities management, gas and central heating, healthcare, office machinery and courier, transport and retail logistics. Whilst they come from diverse industries and have a range of challenges and goals, our customers are nevertheless similar in that they are all striving to do more with their data. They are all at varying points along a path, which we have called the data maturity journey, the end point of which is where we find the most highly evolved organisations that are getting the maximum measurable value from their data, in terms of improved business performance.
Stage one: I have data
The first stage along the journey is to collect and store the data in some way. This is ubiquitous. Field service organisations were amongst the first business users of mobile data, and most create and store vast quantities. But the type of data we store has changed dramatically. It used to be structured data, typically engineers reporting on when they’d started work, arrived at a job, or which consumables needed re-ordering. Now that data is cheap, we’re seeing more unstructured data being captured as well, such as feedback forms, customer service voice calls, pictures and videos of faulty machinery, tweets from happy (or unhappy) customers. As well as this, the data available from the devices themselves, the internet of things, now makes a huge contribution to the amount of data field service companies collect.
Stage two: I can see my data
All this data is potentially valuable, and many organisations have responded by building ‘data lakes’: places to keep all of these different types of structured and unstructured data. But just having your data safely stored isn’t enough. It’s not adding any value unless you can view it, which is the second stage of our maturity model. Again, this isn’t new. Organisations have been doing this for a long time, and there’s a wealth of tools available that enable companies to create reports that can give them lots of useful information to drive the business. Many businesses are managed successfully from data driven KPI dashboards, balanced scorecards and financial reports. However, these reports are designed by humans. You’re relying on them to know both what data will be of value and, how to extract it from the underlying source. Often this requires two different types of person. The first, someone who understands the problem, and the second, someone who understands the data domain and can create the report. This creates a delay. It’s frustrating to request a new or altered report, to be told that a ticket has been raised in your IT department, and you’ll be able to get that report in a few weeks, or even months.
Stage three: I can spot patterns in my data
To address this, many reporting tools have developed user-configurable reporting, to enable the field operations expert to create meaningful reports, quickly and without relying on developers. Once we can easily visualise our data, we can start to see the possibilities of stage three, which is when we are able to spot patterns.
Humans are superb at pattern matching – it was key component in our evolution and survival as a species – but we can only process patterns we can see, which is an issue if the key insight or link isn’t included in the report. Sometimes you need to take a step back to look at the bigger picture so you can spot what is really going on. There are many examples in nature, from migrating birds, to ant colonies and bee hives, where individual components, each following simple rules, combine together to produce complex behavior. This is where computers come in. They, like us, are fantastic at recognising patterns, but have the advantage that they are good at doing simple things many times over, and they can analyse greater volumes and more dimensions of data than humans.
Machine learning excels at this sort of analysis and, with appropriate training data, based on benchmarks from similar companies or industries, you can gain insights that would have been impossible at the previous stage of the journey.
Stage four: I can interpret the patterns in my data
At stage three, we’ve established a mechanism to find patterns we wouldn’t have spotted on our own. However we’re still reliant on our human experts to know what to do with our new found insights, so we can act on them and deliver real financial value to the business. The fourth stage of maturity is to get the Artificial Intelligence (AI) in the system to start making recommendations based on the patterns identified in stage three. By using benchmark data from other similar organisations, AI can recommend a course of action based on the data’s representation of your workforce.
Stage five: I can drive continuous improvement directly from my data
Cloud-native unicorn companies have used this type of automatic data analysis to drive continuous improvement in areas such as the following:
• Netflix recommendation engines knowing what box set you’d likely want to watch next
• Uber deciding where to send idle drivers in anticipation of surge requests
• Sentiment analysis on Twitter predicting whether content is likely to go viral, and generating extra capacity to handle the expected influx of traffic
To these companies, this comes naturally. They have been ‘born in the cloud’ meaning they were never encumbered by internal IT departments that needed months of lead time to provision a new server. If a unicorn company wants to run a report or write a new tool to learn about their product or customer base, they use the data collection, analysis and machine learning tools available in the cloud to do so instantly. They are able to respond to discoveries in their data in near real time, learning as they go, in a virtuous feedback loop.
Data maturity in field service
Field service, once at the forefront of using data to drive improvement, can adopt some of the practices that have made the B2C unicorns so successful. For example, predicting when machines are going to fail, based on data sets from the machines themselves, or collecting recordings of phone calls into service desks, and end-of-day reports from engineers. All of this can be used to feed an automated continuous improvement cycle.
The next few years are full of exciting possibilities for companies looking to succeed by using their data more effectively. Field service organisations must reflect on how they can move to the next stage of the data maturity journey. With all of the machine learning and cloud technologies now available, if we can ‘be more unicorn’, we can all become fully evolved and achieve continuous business improvement through maximising the value of our field based data.