We interviewed Onno Pistorius, the founder and director of ClearPredictions.com, a young start-up from 2015 as well Morphis, a company he started in 2005. We got to know that Onno Pistorius is not only an enthusiastic entrepreneur but an experienced leader in the analytics field and a self-taught developer.
Telling us about himself, he told us, “I did my studies in mechanical engineering and have been fascinated by the possibilities and power of computer software and hence self-taught myself to be a developer to design products that provide a user friendly and easy way of doing difficult processes. As a director, I have been in the lead of the development of various products like workflow management, business rule engines and call center software for Morphis. Using Morphis’ software, employees can easily and effectively handle customer contacts and underlying workflow processes. Morphis was nominated for years in a row for the Deloitte Technology Fast50 and ranked number 38 on the Deloitte Technology Fast 500 EMEA 2010, a ranking of the 500 fastest growing technology companies in EMEA. Morphis was nominated for the category Emerging Entrepreneur in the “Ernst & Young Entrepreneur Of The Year 2010” and received the “FD Gazellen Award” as the fastest growing ICT Company in the region.”
AIM: How did you start your career in it? How has the journey been so far?
OP: I started my career designing industrial robots and innovative, custom designed packaging machines as a mechanical construction engineer. After years of work in that area, I went on to develop ProcessRunner, which is the software suite of Morphis.
I love to make complex things as simple as possible. The journey has been fabulous so far. Since, a couple of years, I have been fascinated with the thought of using predictive capabilities on data – but in a new way. It should no longer be needed to hire expensive teams of data scientists to make use of predictive analytics. Managers and business owners define their strategy mostly looking backwards, using reports, to analyse questions like “what happened?” in their businesses. Predictive analytics can help to provides answers on “what will happen?”. The aim of Clear Predictions here is to provide an easy to use predictive analytics platform for all kinds of businesses and for all kinds of users.
AIM: How do you define Predictive Analytics and how is it related to machine learning, big data and data mining?
OP: “Predictive analytics is the use of data, statistical algorithms and machine-learning techniques to identify the likelihood of future outcomes based on historical data. Bear in mind that, no statistical algorithm can predict the future with 100% certainty; it is based on probabilities.
On the other hand, Machine learning is a method of data analysis that automates predictive model building. It automatically learns to make accurate predictions based on past observations. Big Data is a broad term for data sets so large or complex that traditional data processing applications are inadequate (High Volume, Velocity and Variety). By the way, a CRM system is not big data. Data-mining is carried out by a person, on a particular data set, to get a better understanding of the data.”
AIM: What factors, according to you, have fueled the rise of Predictive Analytics in recent years?
OP: According to IBM, 80% of all the data in the world has been created over the last 2 years alone! This is an amazing trend. The growing volumes of data makes it more interesting to produce valuable information out of it. Besides this trend in data growth, the tougher economic conditions and a need for competitive differentiation are reasons many companies look into predictive analytics now. The goal is to go beyond descriptive statistics (reporting) on “what has happened” to providing best assessment on what will happen in future. The result is improved decision making and predictions that lead to more effective actions.
Analyzing historic data to predict future events enables decisions that can give companies this competitive advantage. These decisions depend on analyzing at a speed, volume, and complexity that is often too great for humans. This would require easy-to-use software and faster and cheaper computers of course.
AIM: Would you like to share any projects that have you been working on this year?
OP: “We have started many projects, one of them is at the utility company Qurrent in Amsterdam. As Mr Slieker, CEO of this company, states: “The ClearPredictions platform is providing us beautiful actionable predictions. It tells us which customer is sensitive to our up-sell campaigns, but also which customer is about to churn. My employees operate the tooling themselves. There is no need any more to hire expensive data scientists nor have long running projects to make use of big data technology.”
We are doing projects for the marketing directors of various retailer companies. Another very interesting pilot we recently started at a leading Dutch Hospital, is about using ClearPredictions’ machine learning technology for an early stage recognition of patients having a rare disease. I am very excited to use our platform in order to make the decision making process in medical care more efficient.
AIM: Why/how are they interesting for you?
OP: Those projects are tremendously interesting for us because we get a lot of high value feedback from them. We honor all thoughtful requests for product enhancements by putting them on our product development roadmap. The high value requests will always to be implemented within a couple of weeks, using our agile development methodology.
AIM: Which industries are placing their bets (i.e. investing heavily) in predictive analytics?
OP: There is a saying amongst marketing managers: “50 percent of my marketing budget was spend well, I only don’t know which 50 percent”. I find this amusing in an era where smart predictive marketing software has become available! The telecom industry is active in predicting churn (customers that tend to leave) for years now. But many other verticals, like assurance, banking, utilities, travel, aviation maintenance and retail, are just getting aware of the possibilities and benefits of the predictive analytics technology.
In order to obtain differentiation towards their competitors, companies are now getting up to speed in realizing the importance of data analytics and have started to investigate.
AIM: What are the biggest areas of opportunity in predictive analytics?
OP: Big data is a megatrend that touches so many aspects of our interactions in life – from the Internet of Things (think about all the data generated by your smart phone, like the GPS positioning feature) and content analytics (so Facebook can present the most interesting updates especially for you) to customer satisfaction (think about recommendations on flipkart.com: customers who bought this, also bought that). What really is going to make predictive analytics go mainstream is the ability to connect not just with data scientists and technologists but with business people. And absolutely one of the key features to that is making use of this technology plain easy. De-mystify the concept of machine learning!
AIM: What challenges do you see for young entrepreneurs starting out their own firms in Analytics? Any words of wisdom for them?
OP: Entrepreneurs in the area of predictive analytics face the fact that they operate on the front wave of technology. One of the mistakes young entrepreneurs make is to release their solution too late. Reid Hoffman, founder LinkedIn, has made a nice statement on this topic: “If you are not embarrassed by the first version of your product, you’ve launched too late.”
So: in the early phase of your company: go out there! Talk to people, visit companies and be an extrovert from the beginning. This will get you all the inputs you need from talking to customers and enable you to design a product that is really needed and that will be your mantra to success.
AIM: What are your top predictions for predictive analytics in 2016-2017?
OP: “Market size of predictive analytics software in 2016-17 alone is over 5 billion dollars (quoting Forrester). This opportunity origins from the needs of operational managers and C-level leaders who have allocated a significant amount of their budget for big data activities in the coming years.
The last 10 years, the paradigm shift was to bring paper documents, HR dossiers and off-line project administration from the office desks into the cloud. The next 10 years, the big shift in mindset will be to find -and value- the trends in the tremendous amount of data we create. Using predictive analytics, hidden trends and correlations will be made visible and valued.”
Try deep learning using MATLAB