Analytical Platforms of Choice Header

Analytical platforms of choice

Big data…a buzzword we’re bound to encounter almost everywhere. Chances are, you’ve probably had a discussion about expanding your use of big data in your company already. Business Analytics is one of the fastest growing markets. Gartner estimates that 75% of businesses across all industry sectors will invest in analytics in the coming years. Having the ambition to adopt analytics is one thing, implementing analytics software is something else. The first question companies must ask themselves is why. ‘Why do we need to to use analytics?’

Once answered, the remaining question is which technology you want to use to fulfil this purpose. In this article, ORTEC Consulting’s Daan Noorlander (Managing Consultant Solutions & Technology) and Frans van Helden (Managing Consultant Business Consulting) explain how you can tackle these questions and how our approach to successful analytics implementation works.

ORTEC Consulting combines different software platforms to address the different types of problems that customers may have. We see a strong trend in the market to configure and build applications and tools on a strong proven software platform, rather than buying single tools for each purpose. But before we take a deep dive into the question of software, let’s explore the different kinds of analytics and what each has to offer.

Assessing analytics maturity

As we mentioned earlier, in order to integrate analytical platforms into your company, the first step is to define your purpose. This purpose can be related to different types of data-analysis.

The Analytics Maturity Curve for Supply Chain Professionals

In the well-known Davenport maturity curve (see above) these types are presented as stages, but we typically perceive them as different types of analytics, with a purpose and value of their own. The first type is: exploratory analytics, a type of analytics that lets you ‘explore’ your data, mainly by visualizing it in different kinds of graphs and applying simple statistics in order to find relationships within the data. This exploratory phase helps you identify trends and summarize the main characteristics of your data. Descriptive analytics is of a more quantitative nature. It helps you draw summaries and extract metrics from your data samples. Together with exploratory analytics, they form the basis of almost every quantitative form of data analysis. The next stage is predictive analytics, where data is used to detect trends and predict the future using regressions, or complex self-learning algorithms. These predictions are used in a wide variety of applications; from operational planning to strategical decision making. The last stage is prescriptive analytics. This type of advanced analytics takes all the other stages into account and enables you to create models to support decision making at various levels. Usually, prescriptive analytics puts algorithms and heuristics to use in order to find the best course of action.


Choosing an analytics software platform

The four analytical stages discussed above have different purposes and are therefore not easily found in one single software package. When you’re choosing a toolset, you need to take a few things into account. Most importantly, you need to ask yourself if the software is suitable to meet your purpose and you need to determine if it will allow you to deliver the value you’re looking for in a short amount of time. You also need to think about:
  1. User-friendliness
  2. Ease of understanding - how easy is it to explain the outcomes to customers/business users? 
  3. Cost - how much are you able to spend on license fees and how many people will use it?
  4. How easy it is to deploy and what level of expertise you need to develop
  5. Support and maintenance costs

Our Approach

Once we address the requirements discussed above, we can effective choose a platform to tackle the challenge at hand. ORTEC Consulting works with a small number of platforms that already cover 80% of the analytical domain.  This ‘starting kit’ selection of software applications consists of Spotfire, R and AIMMS. These platforms are optionally provisioned in Azure (a Microsoft platform for cloud computing), and deployed via the ORTEC Big Data Portal and the native AIMMS PRO deployment platform.  

ORTEC Platforms of choice

Tibco Spotfire

For exploratory analytics, Tibco’s Spotfire is one of the top-notch applications in the market. Its main advantage is the relative ease of building a dashboard with crystal-clear design. Spotfire allows you to build an insightful dashboard within a couple of hours, and easily customize it afterwards. Basic R-code can be implemented as well. Spotfire’s main limitation is the lack of advanced statistics and optimization, which can be found in other applications.

Descriptive analytics can, depending on the level of detail needed, be carried out using both Spotfire and R. While Spotfire is powerful for data visualization, R is a very powerful statistics tool.

R

Basic statistics are easily applied in R, while more advanced methods may take more skills. Since R is open source software, the use of it is free. The major drawback of using R is the lack of a proper interface. The output is usually sent to other (visualization) software. R is also the most used tool for predictive analytics. All known predictive models and techniques are already built and available as independent packages, which makes it very easy to apply. Next to that, R is the worldwide standard for applying (advanced) analytics, which results in having loads of documentation and users in every sector all over the world.

AIMMS

For Prescriptive analytics, our platform of choice is AIMMS. AIMMS is one of the best optimization platforms available. Mathematical models are easily set-up and solved using powerful solvers. Furthermore, little programming knowledge is required. With some additional experience, an advanced GUI can be built to enable decision-making. The programming environment also offers quite some predictive functionality, and is valuable in displaying and processing multi-dimensional data. The real core competence of the system, however, is the ability to translate the data towards mathematical optimization, linking to well-known solvers such as IBM’s CPLEX. 


Summary


The platforms listed above provide a very strong kit to start an analytics department. But it’s not necessarily true that this is all you will ever need to deploy analytics. Sometimes you need custom-made software or specific tools. However, this overview should be helpful for companies that want to start applying analytics but don’t know how to get started. Hundreds of ORTEC customers have been successful with this approach and we’d love to tell you more about it!

Would you need to know more? Do you want to see how these platforms add value? Get in touch – we’ll be happy discover the value with you in one of our prototyping workshops around these platforms.

Let's Get in Touch and Share Thoughts

We would love to join you in your journey toward data-driven decision making.
Let’s get in touch and share thoughts!