SAP Analytics Cloud #1: Giving predictive power to the people

13 July 2017

Peter Douglas

Peter Douglas

Senior Consultant

Simple to use predictive tools which are embedded in SAP Analytics Cloud (SAC) make for a better and more valuable user experience.  Peter Douglas, Data Scientist, explores the predictive possibilities for the masses.

A year and a half is a long time in the tech industry - SAP’s Analytics Cloud product has steadily grown in features and capabilities since it was launched at TechEd 2015. In particular, predictive analytics has become properly embedded within the product and the addition of R integration has opened it up to more sophisticated and valuable data science applications. Now business users can reap the rewards of automated insight generation without any needlessly complicated setup.
 

The long and winding roadmap

It sometimes feels like an uphill struggle getting predictive analytics off the long-term roadmap and into the hands of business users. The (mis)conception is that it is hard and mysterious and hence cannot be trusted. This viewpoint continues to pervade the business world. The successes of the likes of Uber, Amazon and Google in their application of data science demonstrate its disruptive power.
 
Up to now the solutions offered by SAP in predictive and advanced analytics have been stand-alone and geared more towards business data analysts and data scientists. As a result of this it has been hard to make a strong business case for implementing an enterprise IT predictive solution, particularly when the necessary infrastructure and integration of cleansed data sources has not been addressed first . This is often the reason why “predictive” stubbornly refuses to budge from being a long-term roadmap goal.
 

Integrated predictive tools make for a stronger business case

My first impressions of the predictive capabilities in SAC are that SAP have made a concerted effort to make this technology as user-friendly as possible and to reduce the “time to value” in terms of using enterprise data to derive actionable insight. By embedding predictive tools in analytics workflows, a user can now seamlessly apply sophisticated algorithms without having to switch to a completely different solution or configure the parameters of the algorithm to get a trustworthy result.
 

Can I trust this tool?

When discussing SAC with a client a few weeks ago, one of the main concerns they had was with the trustworthiness of the predictive results. Before I go further on this point, let’s first look at what I mean when I say “predictive” capabilities in this product. Currently, SAC offers the following tools that can be classed as predictive or advanced analytics:
 
  • Smart Discovery & Simulation
  • Time-series forecasting
  • R visualisations & algorithms.
 
On top of the list above, other functionality is planned: “Smart Groupings” and “Smart Insights”. The former is where groupings of data values across several measures are identified automatically and the latter where data points selected by a user  trigger “smart explanation” pop ups.
 

R visualisations & algorithms

Going into more detail and working backwards on the list above, I’ll start with R visualisations. The ability to integrate R code with SAC is similar to the functionality offered in SAP Predictive Analytics (SPA), and this means that more skilled users can create custom extensions that exploit the wide range of algorithms in the open-source “R” data science tool. Business users can then apply these algorithms within analytics workflows without any knowledge of R.
 
This then opens up SAC to endless possibilities beyond the “out of the box” tools that are delivered as standard. But what if you don’t have the skills in-house to create R extensions and you don’t wish to request the assistance of one of our team of charming R experts?
 

Time-series forecasting

Next we move to time-series forecasting. This is fully integrated, such that a user can create a time-series plot on some data and generate a “quick” or “advanced” forecast via a single click. The statistical forecast is based on the historic data feeding the plot and utilises the SAP Cloud Platform’s predictive algorithms. A user can select the number of forecast periods in the output and also specify additional input variables to improve the accuracy of the “advanced” forecast.
 
Importantly, going back to the earlier point on “trustworthiness”, the forecast shows ‘confidence bounds’ to give the user a simple representation of the uncertainty in the result. So, rather than just giving an output, it also guides the business user on the limits of accuracy in the output. If a forecast was to be used to create, for example, financial estimates, then the ‘confidence bounds’ would be useful to quantify limits on such estimates.
 

Smart Discovery

Finally, and probably most importantly, there is “Smart Discovery” which includes the capability to generate simulations. This is really where the predictive elements of SAC demonstrate their true value to the non-expert.
 
From an analytics “Story” created via a data source, such as another SAP system, a flat file, etc. a Smart Discovery can be created by a single click. After selecting the measure to explore, for example, revenue or margin, a host of insight is automatically generated and presented to the user as visualisations and natural language. To try this out, I used a familiar data set from a demo I built last year. This offered a good opportunity to compare SPA and SAC “head to head”.
 
Here you can find out how SAC faired when it was put head to head SPA with a real use case.  
 

Tools that work together

In summary, the predictive tools currently available in SAC constitute a solid start for SAP in their efforts to make this product a serious contender in the already crowded predictive and advanced analytics space. By embedding predictive features within analytics workflows, SAP have enabled business users with no skills in the application of data science to gain valuable insight using data from their line of business.
 
Rather than just taking my word for it, I would urge you to try it out for yourself either through the free trial demo (do a Google search for “SAP Analytics Cloud trial”) which has limited predictive functionality available, or by requesting a demo from our team to get a richer understanding of what there is to offer to your business. Whatever you do next, enjoy discovering predictive in the cloud!

 

About the author

Peter Douglas

Senior Consultant

Focused on helping businesses gain valuable insight from their enterprise data through the delivery of solutions in predictive & advanced analytics.

Since arriving at Bluefin, Peter has had the pleasure of being part of challenging and rewarding projects across the Consumer Business, Manufacturing and Pharmaceuticals industries, for some well-known global names. His primary focus is working with clients to understand and exploit the power of emerging capabilities in advanced analytics and apply this to their business needs. His skills and hands-on experience span a range of SAP technologies, such as HANA / BW, Predictive TPM and CRM.

Peter’s passion for all things “Data Science” manifests itself in championing the predictive & advanced analytics ecosystem of SAP tools as a great opportunity for clients to tap into the power of their data to answer targeted business questions with accuracy and insight.

His previous life as a Quantum Physicist and PhD Grad has left Peter with old habits: in everything he does he loves testing theories and experimenting with new ideas and technology. He’s also addicted to continually learning new things.

Outside of work Peter enjoys life in London and, when he can, explores the World.

Bluefin and SAP S/4HANA - welcome to the one horse race