Product
How To
Best Practice
BI and Tech
Embedded BI
Industry Insights
Analytics

5 Predictive Analytics use cases

A predictive analytics tool is essential for businesses wanting to really put their data to use. But better still, predictive analytics tools can unlock previously unknown insights to engineer real-world change and boost business growth.

Let’s look at how Pi Analytics can be applied to data collected by various industries to generate positive results.

Healthcare

We – together with our partners OCF Data – worked with data from the NHS to study trends in radiology.

We took a large amount of data about x-rays and fed it into Pi Analytics. This information included everything from the patients’ ages, gender, how they entered the hospital (i.e. on foot, in a wheelchair etc.), and more.

However, what we were aiming to measure is the failures of x-rays, where the patient has to be subjected to a second x-ray (and therefore a double dose of radiation).

In this scenario, Pi Analytics told us that modality (the way the patient entered the hospital) was the main factor in why x-rays had failed the first time around.

Armed with this information, the hospital could then assess how they could x-ray patients better who arrived in a wheelchair or on crutches, for example, to lower the number of x-rays having to be retaken.

Healthcare Analytics showing reject analysis of x-rays.

Education

Students dropping out of college and university is a constant problem for those institutions. After all, when a student completes their course – especially attaining a good grade – the college or university has a better chance of ranking higher in league tables and is in a better financial position.

When Pi Analytics was fed a substantial amount of education data, it was able to determine the main signs that a student is becoming disenchanted with their course. These factors included attendance, whether they had a library card, how involved their parents were, and more.

With this information being made accessible, colleges and universities can target the students that are displaying these signs, and stage interventions to increase their chances of continuing with their courses.

Insurance

The Insurance sector is no stranger to analytics tools, with insurers having to gauge whether applicants are appropriate to insure or, alternatively, whether they’re not.

Pi Analytics can go over and above what older analytics tools could offer the industry. Its intuitive user interface means domain experts don’t have to rely on data scientists to create the models, they can just self-serve their own insights.

Our predictive analytics tool can help your users to identify how best to shorten the claims cycle, manage risk, and improve win rates, as well as achieving consistency in pricing and being able to identify trends and patterns to eliminate fraud.

Finance

We were previously asked to create an analytical model over a data set that comprised of 16 million loans to discover any loans that were defaulting.

A small subset of people who matched that criteria were found – just 90 people out of 16 million. Within this group, all these people were found to be women, who were over 60, and had a high credit rating.

The domain experts quickly realised that these women were likely recently widowed. Their teams were immediately informed not to chase these people hard for payments. Instead, a new approach was required, given the sensitivity of the situation.

Finance Dashboard example showing invoices overdue

Logistics

Analytical insights also bode well for the Logistics industry. Pi Analytics could predict potential disruptions in the supply chain, to reduce the uncertainty in lead time, which is always a factor when shipping overseas.

When fed with data relating to traffic patterns, weather conditions, and port behaviour, Pi Analytics could more accurately predict times of arrival, which achieves greater levels of transparency for customers, and increased satisfaction when products arrive on time.

If you’re ready to see what Pi Analytics can do in your industry, get in touch to organise a trial run with your data, and see what undiscovered insights become known.

Written by Matt

Matt is Panintelligence’s Digital Marketing Executive, who’s responsible for the SEO and content strategy of the company website. He previously spent four years working at a leading digital agency in Leeds. When Matt’s not optimising the Panintelligence website, he can be found playing video games, managing his YouTube channel, or out on a walk with a guaranteed pub stop.