A few weeks ago the highly anticipated Gartner Magic Quadrant reports for Business Intelligence and Analytics Platforms and Advanced Analytics were released. Each year I anxiously search the internet trying to get a first peek of the final results. I cherish reading every single word of these findings to see how they compare to my own experiences and market analysis. The past several years I shared my thoughts on those rankings in a colorful review of the review per se. Those sassy and honest posts have been top viewed, reader favorites!
This year, I will share thoughts on the rankings with you in a much more reserved manner since I am currently not independent. I will also explore the state of the market along with a few predictions that I feel are inescapable in the midst of mass decentralization and impending data growth explosion.
Understanding Gartner Magic Quadrant Context
Many vendors ranked on this report will highlight Magic Quadrant results to customers and prospects. Savvy marketing minds craft messaging campaigns to highlight identified strengths and exploit competitor weaknesses to spin findings in the best light possible. Thus the results of these reports are often taken out of context.
Marketing gurus bake Magic Quadrant results and stories into the core sales pitch, overview decks, proposals, web sites, email signatures, collateral…you name it. That is what marketing is all about – right! Framing the perception of a good or service in the mind of a customer in every touch and interaction.
Even though Gartner includes a dedicated section on context along with guidance on how to properly reference the Magic Quadrant report along with other Gartner reports, most of the time customers and BI professionals themselves are unaware of that section. Few people look past the well-known quadrant graph to read the 90-page summary report on market trends, changes made to the evaluation process, criteria and vendors not shown. Did you?
In the context section, Gartner states that they have “defined the BI and analytics market broadly. We include a variety of products that span a range of buyers and use cases“. Last year I mentioned the ranking compares “apples to oranges”. There are different offerings serving completely different audiences and needs shown side-by-side on the exact same chart i.e. an ETL tool and a visualization tool.
There are also completely different scopes of solution offerings shown on the same chart. For example, SAP, IBM, Microsoft and other vendors have numerous components that get evaluated from master data management, metadata management, data modeling, data quality services, ETL, OLAP, ad-hoc reporting, operational reporting, search, data discovery, dashboards, alerting, collaboration, and so on. Alteryx, Tableau, Datawatch and others may only submit one component.
My bottom line message to you is not intended to discount the success of SAS, SAP, Tableau, Qlik, Microsoft, and others in the Leaders Quadrant. I want to be sure you actually understand the context of this report as a trusted BI advisor and professional. If you evaluate vendors, go the extra mile to read through the entire document. There is a lot of excellent insight and forward looking tips within that report.
Random Notes of Interest
Due to my current status, I can’t share all the forward looking market strategy puzzle pieces that I took away from the 2015 Magic Quadrant report. However, I will share a couple interesting points.
This year Gartner seemed to make more changes to the evaluation criteria than I recall seeing in past reports. They mentioned dropping Microsoft Office integration as a stand-alone critical capability. It sounds like the evaluators moved it to a lower scoring, sub criteria section under IT Developed Reporting and Dashboards (Produce). In my opinion, there could be a myriad of reasons for doing this in light of Google Docs and other Office-like solutions in the modern, mobile, browser-based market.
Gartner also mentioned “shifting requirements from IT to the business”. I have to imagine that change also affected scoring. I am guessing it might be one of the reasons why all traditional vendors shifted down this year.
The notes on seeing more embedded analytics into applications, smart analytics and self-service data preparation does reflect industry trends that I have seen the past year.
I also felt the points on needed future big data scalability and increasing requirements for streaming and real-time analytics were key. I have been ramping up on the new big data, collect and ingest everything patterns with Lamda Architecture, Hadoop, Kafka, Event Hubs, Azure Data Factory, Apache Storm, Azure Streaming Analytics and real-time dashboard development with the Microsoft Power BI Preview.
Predicting a Return to Balanced Approaches
The current shift away from centralized to decentralized analytic approaches was clearly the take-away theme in my mind in this year’s report. Why? Gartner cites “better business value”. In my opinion, the business appreciates quick, build-it-yourself reports. There seems to be less concern if reporting is done properly. Data discovery vendors are selling direct to naïve business users. The reporting tools today are easier than ever to use but the business users don’t know what they don’t know. What pains will need to be experienced before we return back to a balanced approach?
Personally I have seen both success and failure with the decentralized approach and the top data discovery tools. All of these players ding Excel reporting but literally drive you to the same end-state with a more attractive user interface. Data discovery tool failure looks a lot like Excel/Access/VBA or even Operational Data Store (ODS) data messes simply with your data discovery tool of choice. You know you have a data mess when:
- everyone has their own version of a calendar, person or other entity that should be shared
- if you can only report with today’s organizational structure or values
- if reports and data sources built by everyone use their own naming conventions
- there is wide open access to all reports with no role-based permissions applied
- copies of similar reports have slightly different fields, formulas and totally different values
- numerous copies of connections are used for the same data sources
- data that does not refresh ends up being a personal file that was accidentally published instead of the connection to the real data source
- too many data sources are trying to be mashed-up or blended
- business logic in front-end scripts become impossible to sort through and easily get out of sync
- reports have numerous formulas with almost the same name that compute the same value
- reports can’t be easily updated without breaking a formula
- if a report is accidentally overwritten, there is no means to roll back/recover from that error
- if data is incorrect, there is no means to easily reload or update it to apply corrections
I could go on and on. I personally feel it is scary that Gartner cites “in 2016, less than 10% of self-service business intelligence initiatives will be governed”. I am all for empowering the business responsibly. Decision making with reports that have no governance can be damaging, potentially illegal or non-compliant with needed requirements/safeguards.
There are proven designs for accurately reporting data over time with slowly changing dimensions and other data warehousing patterns that are being ignored in the get reports built fast, decentralized, data discovery world. These patterns are still being applied in the modern Hadoop, streaming, IoT, and big data world. It will never be old-school to report correct numbers and accurately compare historical state with current state! If you can only report with today’s organizational structure, you may be in big trouble. Needless to say, it is not easy or cheap to clean up an ungoverned data mess. With the exponential growth of data and new design patterns for storing all kinds of structured and unstructured analytic data, I feel the data mess issues will only get worse if governance continues to be an afterthought.
How the Vendors Fared
In my reserved personal opinion, I saw no surprises in 2015 except the shift of all traditional vendors down a notch.
- Top 3 Leaders/Visionaries: SAS, SAP and IBM
- Top 3 Leaders/Ability to Execute: Tableau, Qlik and Microsoft
I thought it was interesting that a few vendors including IBM and Microstrategy were given hints in this year’s report to improve various areas to retain position in the Leaders quadrant. SAS excelled in vision with Visual Analytics though I rarely if ever run into them. SAP also fared well with the SAP Lumira and KXEN integrations.
Although I gave Tableau a rough time after TCC in my ho-hum post, it seems like they held back a bit in that key note last year. I am seeing notable new v9 features that I did not see while at TCC. Qlik Sense, much like the current Microsoft Power BI Preview, was a completely reimagined user experience. Qlik just happened to effectively brand and position their new user experience with a warm, fuzzy name “Sense” and a themed “Natural Analytics” campaign. I reviewed it last summer and thought storytelling was fantastic. Although Qlik Sense is a v1 offering, I do respect both the vision and governance aspects in their analytic approach. It is similar to the Microsoft BI approach.
Speaking of Microsoft BI, I was relieved to see that “Microsoft’s strong product vision and future road map” for Power BI were well-received along with governance capabilities and other positive changes in both leadership and offerings. That is good news for Microsoft BI fans going into 2015.
Datawatch made a strong debut and will be one to watch this coming year. Atleryx fared well again this year. Pyramid Analytics moved up a bit while both Birst and Logi held ground. DOMO and SiSense, who I do run into from time to time, did not make the top 24 vendor cut. I did find it intriguing that DOMO is going to be less secretive starting in April 2015.
Looking forward to 2016
With rapid development cycles, market shifts, newcomers constantly entering the game, big data analytics challenging and altering data gravity, I can’t wait to see how next year sorts out. I still can’t believe how much the BI world has radically changed in the past five years.