From my understanding of the marketing of predictive analytics in SaaS products I believe that more and more business problems will be tackled in a standardized way, i.e. you want to know the risk of fraud in this next credit card payment is? Just push the credit card number to an API and have it assessed. You want to know the future lifetime value of certain customer so that you can optimize your marketing expenses on this person. Push the the customer to another service that will provide the information needed.
Today, many companies still approach questions like the above by building a data warehouse, a team of highly specialized data scientists (i.e. predictive analytics professionals) and invest time and money and time and money again. Surprisingly this is often worth it. The technology behind all this is so powerful that you can easily invest vast amounts of money and still yield a positive return.
But this will change. And it has already started changing. And the speed of change is going to accelerate. Watson, Einstein and so on will put so much pressure on companies to move fast that they will have to take new approaches. Building a predictive analytics model and maintaining it consumes a lot of your resources. And as soon as any of the processes nearby are changed you will need to invest into the predictive model again.
So, what I expect is that we will see many many new predictive analytics services in the near future that companies will start to embrace instead of trying to solve every problem on your one which has never been an efficient approach. This way companies will be able to make use of much more predictive models which will ultimately take them to a higher level of competitiveness.
So, do the predictive analytics professionals have to walk home? No, not at all! There will be two common paths. One is to join a company that develops such services and to become a true expert in what you do. From my experience the lack of excellence in project based work is one of the big drawbacks in data science work. Developing a single product offers something completely different. And the other path will be to manage all the predictive analytics services for a company. This is quite different as you need to have a deep understanding of the data science going on in the services as you will otherwise not be able to make use of the true potential of those services and you also need to have a true understanding of the business domain you are working in. Only when these two conditions apply will you be able to really outsmart your competitors. At Gpredictive we adapted the data artists for these positions rather than data scientist.