In the latest round of 2017 reviews, Gartner released a Magic Quadrant for Data Science Platforms. Results this year were less surprising to me than the BI and Analytics Platforms. Even the name change from Advanced Analytics made sense. Earlier this year, I updated my Advanced Analytics menu category to Data Science on my blog since I was seeing the same market shift in terminology. Let’s take a quick skim through the results.

No Surprises

This year the team that analyzes advanced analytics aka data science platforms completely revisited the criteria allowing a mixture of old and new vendors in. I liked that approach as it can be incredibly difficult for small, innovative startups to get a chance to compete in the big leagues. Like all Magic Quadrant reports, it is important to read the entire report to understand what is actually being scored to not confuse vendor placement with assumptions on solution capabilities. Vendor solution features and functionality are more clearly scored on Gartner Critical Capabilities reports.

Data science platforms are described by Gartner as being “a cohesive software application that offers a mixture of basic building blocks essential for creating all kinds of data science solution, and for incorporating those solutions into business processes, surrounding infrastructure and products. Cohesive means that the application’s basic building blocks are well-integrated into a single platform, that they provide a consistent look and feel, and that the modules are reasonably interoperable in support of an analytics pipeline. An application that is not cohesive — that mostly uses or bundles various packages and libraries — is not considered a data science platform, according to our definition.”

The report went on to describe the types of types of data scientists, stack ranking process and criteria.

IBM a Leader Again

IBM remains a leader this year. Having personally used IBM data science ecosystem solutions SPSS and more recently Data Science Experience (DSX), I can share that they are deep, powerful yet easy to use. I loved the DSX and have been showing it in my predictive presentations. Although DXS was not scored, Gartner noted it helped with IBM’s position on the Completeness of Vision axis.

I did not see much mention of IBM Watson cognitive computing in this report though IBM Watson Analytics was mentioned. In my opinion, IBM Watson cognitive computing is compelling. I felt the REST APIs were easy to learn, integrate and again…they are super powerful.

The note on confusion is fair. I was lost trying to learn about IBM’s ecosystem. There are many SKUs in their portfolio that do seem similar.

SAS also a Leader Again

SAS is another market leader in predictive analytics and has been for as long as I can remember. I learned SAS Enterprise Guide and SAS Enterprise Miner years ago. I have not had the pleasure to learn the Visual Analytics Suite yet. Usually when I tweet about SAS or even IBM for the matter, someone in my network will chime in to comment about open source R and Python.

SAS, IBM and other vendors in this report are all embracing open source.

What I have seen happen is traditional vendors provide better integrated user experiences with open source, scale and support in exchange for premium pricing.

SAS has not briefed me yet. They have sponsored me to attend SAS Global Forum in April where I will be getting more information directly from SAS product teams and hands-on evaluations with the latest offerings. At that time, I will write more about them.


Learn Data MiningRapidMiner is great lower cost data science solution that I have also been showcasing for years in my predictive analytics classes. Data Mining for the Masses is a lovely guide to learning this specific offering. I recommend it to my students for additional hands-on exercises.

I felt the strengths and cautions were fair. I personally found RapidMiner data flow input output connectors to be a bit confusing to learn at first. From what I have seen, they do continue to get better and better each year.


Seeing Knime as a leader is also not surprising. I was recommended to Knime by a respected enterprise client of mine back in 2012. I downloaded it, gave it a whirl and found it to have all the essentials that I needed. In data science, there are a lot of low cost and free offerings. Knime is one of them but it is also deeper than most. Thus it has a strong placement here and a healthy community.

I am now starting to see Knime more at business analyst community events. They seem to have stepped up marketing as of 2016 and it shows.


I was pleasantly surprised to see H2Oai in here! In my December Spark Big Data Analytics webinar, I introduced and demonstrated how to use the H2Oai offering. I also highlighted this little company in my Winter Industry Pulse magazine as a hot startup. Playing hands-on with H2Oai, I found it to be easy to learn and awesome. The offering is really quite cool and super powerful in the mega-hot Spark big data analytics ecosystem.

hot startup

Incorta, another one of my Winter 2017 picks for a hot startup just announced $10M in Series A funding today. I do know a good solution when I see it! Finding these gems is usually the hard part.


I felt Microsoft’s placement was fair and consistent with prior years. In the past, I was simply ecstatic about the acquisition of Revolution Analytics. I am a longtime fan of that offering. Although I haven’t played with Azure ML in a year now, it should be more mature. I am taken back by the continued limitation in integration options to make predictive algorithms easier to operationalize. I had trouble with that myself and routed results to Azure SQL.

The lack of on-premises does limit use and customers. I see on-premises groups using SQL Server R options instead of Azure ML. Azure ML does (or did) have an Excel Add-In that I did not see any mention of in here. Azure ML looks a lot like Seahorse Visual Spark framework that I do enjoy developing on. Azure ML should eventually evolve to be much nicer than it is today. Visual Spark framework


Dataiku had a big year in 2016. I literally saw them almost everywhere I went. A few readers asked me about them so I inquired about a demo of this solution. However, I never got a reply. If I do see them at the Gartner event next week, I’ll see if they would be willing to do a Solution Review. I like the sound of the persona-based design user experiences. I need to see it and try it to comment on it in any detail.

Other Vendors

SAP’s placement and write up might be the only surprise to me. I have not done much with SAP this past year. In the past, I did write about the KXEN solution and Infinite Insight. I was looking forward to seeing what came of that offering. I do like SAP Hana Predictive Analytics Libraries (PAL) for in-database analytics. In general, I find in-database analytics to be easy to operationalize into smart reports and dashboards.

Alpine Data Labs is another solution that I have written about in the past. I liked it but I have not reviewed it in several years. I don’t run into them much these days nor do I see them in the market.

Alteryx fared well. I do like Alteryx’s easy drag-and-drop user experience super much. I wrote about it several times. I have heard mixed reviews on the baked in predictive scale so that comment in the cautions was spot on to what predictive users told me this past year.

I don’t see the other players mentioned in this report often. Several honorable mention vendors that I do indeed see gaining market traction include Amazon, Databricks, DataRobot and Google. Next year should be more interesting as these up and coming players mature.

Right Time for Data Science

In a digital transformation, using collected data for analytics, data science and artificial intelligence becomes the ultimate competitive weapon. I see this market exploding right now. It will just continue to get bigger and bigger. Today CBInsights tweeted this interesting image.

AI Trends

It is the right time, right place to be in the data science space.