Four reasons why getting started with predictive analytics is simpler than you think

We spend a great deal of our time at Smart Vision helping our clients to establish the use of predictive analytics in their business. For many organisations, getting started with predictive analytics can feel like a real departure from more traditional and familiar areas of activity. That said, organisations in almost every industry sector are becoming increasingly aware that to maintain a competitive edge it’s necessary to have detailed customer, product and operational insight; and that data analysis and modelling of organisational data is a required capability and key source of competitive advantage.

In this post I want to talk about why investing in and developing a predictive analytics capability in your organisation is different (and better!) than getting started with other kinds of technology projects with which you may have more experience.

Implementing predictive analytics is less expensive, quicker and lower risk than almost any other kind of technology project

Setting up predictive analytics is very different from more traditional technology products such as implementing a new CRM, financial or other operation system. The mere mention of implementation of new operational systems often sends a shiver down the spine of IT Manager and line of business leaders alike. Such projects are often tarnished with a history of missed deadlines, significant budget overrun and seemingly never-ending creep of scope. This often locks the client and the vendor into a long and uncomfortable enforced relationship.

Predictive analytics is a much more agile set of technologies to establish. You’ll often be able to set up the relevant software & hardware, connect to appropriate data sources and begin work in a few hours or a small number of days; it’s certainly not a question of weeks or months. This agile implementation is painless. Moreover, a quick and simple initial set up in no way compromises or contradicts the support and planning of a longer term enterprise approach.

In my time in the predictive analytics industry I’ve seen many examples of clients being able to set up an analytics project and deliver a significant financial return to their organisation, all within a few short months, and for less than £20K initial investment.

You already have much of the expertise you will need in-house

Another common cause of anxiety from clients is the concern around a skills gap. Do you need to go and hire a PhD-level operational researcher or statistician to do this for you? The simple answer to this is no.

In almost every project I’ve worked on, we’re training and enabling data literate, business-focused people in the organisation to become analysts. You probably already have lots of people who fit this profile in your organisation. These people can quickly develop the analytical skills and understanding required to become very effective and, crucially, practically productive data scientists.

This is good for your business as you keep and develop skills in house, and develop your reputation for investing in your people’s skills. Of course, if you already have people with high level statistical or analytics skills in the business then that’s great, but it is certainly not a prerequisite. It’s the business acumen and understanding that are the more important attributes, I’d suggest. Analytical skills and understanding can be layered on quickly. It is the combination of these two skills sets that will catalyse your transformation towards becoming a data savvy organisation.

Your data will be fragmented, dislocated and of variable quality but your project will still succeed

I have never been involved in any analytics project where the required data existed all in one location, beautifully cleansed and fully validated. Messy data is an unavoidable occupational hazard but is not a barrier to getting started.

The quality, accessibility and range of data is a challenge that has to be addressed to some degree at some point of course. The practical reality is that the very undertaking of implementing predictive analytics moves your organisation towards addressing this area. In fact, the process of starting to answer key organisational questions through the analysis and modelling of data will have multiple effects. Ultimately it will of course enable you to answer those central business questions you want answered, but it will also deliver the following insights:

      • You’ll learn what data is the most valuable when addressing customer insight and operational questions
      • You’ll quickly identify gaps in the data you have available so that you can then decide whether it is possible and practical to fill them
      • You will develop a deep understanding of what an enterprise view of data from a predictive analytics perspective needs to look like

The reality is almost all data, in almost every type of organisation, was not originally collected with analysis in mind (with the exception of research data). You can’t change that. You must work with what you have.

You will need to decide which question to address first and how you will use the answers you get

One of the concerns we often hear is ‘Where do we start, what question do we tackle first?’ The potential applications of predictive analytics in any organisation are going to be numerous. Are you addressing issues of customer profitability, customer retention, the key drivers of satisfaction, operational efficiency, pricing optimisation, product warranty performance or some other issue? The options can be dizzying. The reality is you must start somewhere and you must be specific about your area of application, focused on the detail and the measurement. We often talk about this as ‘setting the exam question’.

You will need to be clear about what exactly you are asking of the data and how you are going to answer it. In particular you must think through how, having answered question set, are you going to change behaviour, actions and policy to deliver different outcomes. You must also agree and test how you will measure the impact and at what level is a changed outcome considered a success.

In many ways it is this last area that asks the most of an organisation and its people. Perhaps ironically, this has very little to do with technology. Using the insights gained from data analysis to change organisational and individual behaviour and decision-making is always a worthwhile and profitable undertaking. The practical aspects of the technology investment, ensuring you have the right skills and having the right data available are not the biggest challenges. The biggest challenge is making sure that you implement your findings, allow changes to behaviour and process to actually happen and then measure their impact, and it’s this that will be the ultimate arbiter of success.

Scroll to Top