Gartner Magic Quadrants – How to make choosing the right data technology a piece of cake

Birthdays and anniversaries come around each year with anticipation and excitement. For most of us, the passing of the years brings a deeper sense of calm as we mature – and save for the challenge of blowing out an increasing number of candles precariously balanced in the icing – annual celebrations take on a reassuring predictability as time marches on.

Not so in the Data technology landscape. For this year, Gartner have unexpectedly changed the recipe of their Business Intelligence & Analytics “Magic Quadrant” – the 2016 version of which was published earlier this week.

Most shockingly, Oracle has been dropped from the 2016 BI Magic Quadrant. This is so profound it warrants labouring the point. The Gartner Business Intelligence Magic Quadrant, which is to many the de facto Market Guide to the Business Intelligence industry, and now one that does not feature Oracle… Have Gartner baked a cake without eggs? Will Lary Ellison be choking on his battenberg?

The 2016 Gartner Business Intelligence and Analytics Magic Quadrant The 2016 Gartner Business Intelligence and Analytics Magic Quadrant

First we need to start with a little background:

Management Information Systems and in particular Business Intelligence solutions have been the ground-zero of the war between Business and IT (see my earlier article: https://www.dataphilos.com/blog/2015/10/15/data-and-the-war-between-business-and-it). In a world where data is the new oil; control over supply, distribution, refinement, productisation and eventually monetisation have been the strategic battlegrounds over budget and resources. In the old world order it was easy, the Business users wanted insight but didn’t have the skills or tools to get information together in a timely fashion. IT would offer solutions, but in doing so either constrained user behaviour to the workflows available from the tools in the market, or genuinely made a best endeavours attempt of understanding the needs of the business only to discover that by the time the project was complete (note – BI projects are never complete) the needs of the business had moved on. In both cases the necessity for the Business to do their own reporting and analysis was solved by the spreadsheet.

The first generation of these systems represented a need simply to digitise the real-world. A system of ‘books and records’ was the silicon son of its paper parent. The obligation to hard-wire solutions to satisfy the needs of accounting departments created a barrier to truly harness the power of the technology in gaining insight from the data that flowed through an organisation. Business Intelligence solutions offered a way past this, by building giant ‘data warehouses’ – organisations could start to find operational efficiencies and competitive advantages by designing their data supply chains. Executives could now be given dashboards and heads of operations could be given reports on their fiefdoms, updated as frequently as they desired – sliced and diced as much as their imagination would allow.

This is largely the world we still live in today, but something remarkable changed a decade or so ago that begun to challenge the status quo.

Business Intelligence solutions were often cited as having failed for running over budget, or failing to provide the insight desired in a timely way. While many of the established vendors (such as Business Objects, Siebel, Hyperion or COGNOS) continued to focus on their Enterprise (read IT-centric) capabilities, a new breed of software company such as Tableau and Qlik focussed on the needs of the business user and gained traction with people who needed solutions to problems which they either didn’t had the time to solve with Excel, or Excel’s limitations meant they had to search for another solution (without telling their colleagues in IT).

The analysts and the marketeers realised that this new subset of user was starting to account for an increasingly significant proportion of the net new spend for BI tools, and they ought to give them a name. The term ‘data discovery’ was coined to describe user-centric and user-driven BI, and by 2006 QlikView was featured as a ‘visionary’ in the Gartner Business Intelligence Magic Quadrant. Over the last decade they have been joined by their rivals such as Tableau, Spotfire, and others. The result being in the last few years that ‘data discovery’ vendors are recognised as ‘leaders’ in the BI industry by Gartner and others; rubbing shoulders with the big industry names such as IBM, SAP, Oracle, and Microsoft.

The birth of internet marketing is largely responsible for another evolution of the data analytics industry. Since the dot.com boom at the turn of the century (and in particular driven by its crash and recovery in the last decade), the need to perform complex analytics on rapidly moving sets of data in order to arbitrage and optimise advertising campaigns or to increase conversion rates of users on website, has given rise to a new breed of user and technology. What we now call Data Science is a broad church, but in whatever industry they hone their craft, Data Scientists need tools to help leverage their time better in creating statistical models that best represent underlying behaviour expressed through data. Very few of the established Business Intelligence or Data Discovery providers could offer this capability, and so in 2014 Gartner broke out a new Magic Quadrant (MQ) for “Advanced Analytics” tools, of which we now see the third iteration.

While the 2016 Advanced Analytics MQ doesn’t hold many surprises (it’s nice to see my friends at Lavastorm finally getting recognition, but seemingly at the expense of Tibco who were dropped); Gartner have turned the world on its head with the publication of the 2016 Business Intelligence and Analytics MQ this week. Rather than re-defining the ‘data discovery’ providers as ‘analytics’ tools and breaking the out into their own Magic Quadrant, Gartner have instead chosen to redefine what constitutes a “Modern BI” solution and instead have ruled out traditional BI requirements from their criteria. We can instead expect a “Market Guide for Enterprise Reporting-Based Platforms”, to be published alter this year, and consequently vendors with products which major in ‘legacy’ requirements such as industrial-grade reporting (what many people still think of Business Intelligence to mean) have been downgraded or simply dropped.

Throughout the 2016 BI MQ mention is made of vendors’ traditional offerings being disregarded in their placement in the Magic Quadrant. SAP scores nothing for Business Objects, IBM nothing for COGNOS, Microsoft nothing for the SQL Server family, and as for Oracle? Nothing. Nada. OBIEE bye bye.

I do agree that the ‘traditional’ BI landscape is now mature and save for Big Data architecture challenges needs have not developed greatly over the last few years. The problem is however that by essentially replacing all the ‘traditional’ BI vendors with their ‘new-age data discovery’ alternatives in the new MQ, Gartner seemingly contradict their ‘bi-modal IT’ system which they championed in 2015 (bi-modal IT is worthy of its own article, but for now read this as both IT-centrally governed and end-user developed BI in parallel. It’s like Quantum Theory for Data Philosophers)!

Better would have been to have broken out the Data Discovery products into a new Magic Quadrant, rather like Advanced Analytics in 2014, or at the very least, to have released the Market Guide to Enterprise Reporting Systems ahead of the Magic Quadrant to avoid any potential confusion. For the challenge as it stands is that while most practitioners will read the BI & Analytics MQ as the “Data Discovery” Magic Quadrant, most buyers will not.

With the new criteria comes a new champion, and it seems that in 2016 most lay-people will recognise Microsoft as the Kings of BI (sadly most people don’t read the MQ, but simply gaze at the Battenberg-cake chart). I would agree with this assessment, but for slightly different reasons to the lay-person and to Gartner (and with significantly less research, I should add)! For me, Microsoft is the only vendor with a complete offering. Power BI represents a functionality rich set of power-user analytics tools, and the SQL Server suite is a mature set of Enterprise reporting solutions. It’s the only vendor that offers all components (although arguably with Qlik’s acquisition of nPrinting in 2015 it also finally offers a true reporting capability). Additionally with Microsoft’s acquisition of Revolution Analytics combined with its own Azure ML platform it is the only vendor that also offers the toolkit to satisfy the Data Scientists amongst us.

Which leads me to my final point. The latest trend I’m noticing in the BI space is for Natural Language based search analytics. Tools such as Thought-Spot (http://www.thoughtspot.com/) trail-blaze with capability that enables the business user to pose questions to the system with ‘Google-esque’ ease and no analyst overhead. However, to me – this represents another new category of BI, and one where all the vendors, both established and start-ups are likely to be fighting over in the next decade. It’s also another area that Microsoft have a solution – with their Cortana Analytics offering. It seems that Microsoft have finally woken up to the opportunity a serious play in the Analytics space offers (having slept on Excel for over 20 years). My only hope is that as Natural Language search-based BI starts to take hold, Gartner recognise this for the distinct category that it is, rather than conflating it into the same criteria as Data Discovery and Business Intelligence before.

Cake, anyone?

Leave a Comment