Why I Can Safely Recommend Tableau to Large Enterprises Now

Growing up/evolving is hard work

Growing up is irritating, boring at times, and full of responsibilities. The need to consider all sides of an argument and not rush to change the world at every step, moving slowly (if required) in order to not break things, and developing empathy are all qualities that are valued in a mature organizational setting, but so difficult to inculcate for the most of us - me included.

But that's the reality of the large enterprise world. Things can be 'cool' and 'hip', but not at the cost of 'stable' and 'scalable'. Brevity and engagement, but not at the cost of clarity and functionality.

But enough with the long preamble. 

Pop Quiz: What does Tableau do, other than visualization?

Chances are, if you have heard about Tableau, you are aware of its visualization capabilities, or the way it makes the analytics experience engaging for non-expert users. There is no denying that Tableau has been a poster child for self-service visual analytics, helping business users break free of draconian, IT-driven report factories. At the Tableau on Tour in London 2016 event, I met users whose careers have skyrocketed because of Tableau – or more accurately, their ability to find problems and opportunities in their business area with Tableau, which made them skilled industry experts highly desired by employers all and sundry.

That's amazing. At this point, if you feel an undying urge to cheer or wooooo, please sit down and hold it in, just for a little while. Yes, that means you too, Raj and Kevin - I see you. (names are not real)

Source: Twitter (https://pbs.twimg.com/media/ClG8lqXWYAA8dr_.jpg:large)

Because while Tableau loves you, visualization is not all Tableau is about.

As analytics customer/user expectations evolve in line with the market, even very large enterprises have started considering and deploying Tableau – not as a departmental solution, but across the enterprise and with IT involvement. Large organizations that come for the visualization and user empowerment story, need scalability, multi-structured data management across multiple sources, and interoperability with diverse enterprise applications to stay. At the event, Francois Ajenstat, vice president of product management, demonstrated that enterprise offerings and mature data management capabilities are very much a part of Tableau's story and vision. The vendor understands that visualization and self-service is no longer rogue or shadow IT, and a seat at the adult "enterprise" table comes with a certain set of expectations.

Visualization is thriving, but the market is getting crowded

Don't get me wrong. Easy and elegant visualization of data was, and will continue to be, the single biggest factor driving analytics into the hands of underserved masses. The attractiveness of this market is evidenced by the launch of visualization solutions from almost every vendor in the market, including mega-vendors that dominate the BI world (IBM with IBM Watson Analytics, SAP with SAP BusinessObjects Lumira, Microsoft with PowerBI), but also from so many others (Salesforce Wave, Amazon Quicksight, and the latest, Google Data Studio 360). Not every vendor has the same strategy though – for example, Amazon QuickSight is heavy on AI/machine learning for a guided experience, while IBM Watson Analytics can use IBM Watson APIs to deliver a cognitive experience. In my recent “Market Landscape: Self-Service Visual Business Intelligence/Analytics, 2016”, we covered no less than 18 vendors making a play at the self-service analytics pie.

Vendor Landscape in the self-service analytics world, not exhaustive

Source: Ovum, company logos from websites

But with the appetite for self-service analytics growing exponentially, the real challenges of managing and harmonizing multiple analytic and data environments are coming to the forefront. Getting analytics to the desk of "everyman" is indeed desirable, but it cannot come at the cost of creating multiple silos that provide an inconsistent view of the business at best, and expose the company to significant operational and compliance risk at worst. Large enterprises with complex data analytics needs and massive distribution requirements must carefully walk the line between a data/analytics free-for-all and a draconian locked-down environment, as neither of these extreme approaches is sustainable. I alluded to the prevalence of this problem in our recent How-To Guide: Enterprise Analytics and Business Intelligence report, and strongly believe that the key to moving from islands of analytics to a connected ecosystem lies in pursuing strategies that focus on data, providing a logical layer for data access, integration, and management. "Who owns the query?", has been a perennial bone of contention, with database vendors and BI tool vendors wanting to be the portal for query and to take charge of pushdown, because data is never all in one place. This is possible today with modularity of query and analytics with the ability to embed, API-based access to data and schema/logic/metadata, processing pushdown to database, and the ability to abstract out a few layers of technology while accessing data. In my opinion, analytic software/methods that deal in data extracts and maintaining redundant disjointed copies of data are counter-productive to this idea.

Tableau sharpening its enterprise hooks with data management

Tableau has got the message loud and clear. With over 42,000 customers and counting, it is increasingly getting into larger deals where IT is involved from the pre-sales stage, and the deployments are massive from day one – not the typical "land and expand" approach that has driven Tableau's growth. Recent client examples that lend credence to this theory include La Poste, McFarlan Smith, Aeria Games, Schneider Electric, and Burberry, to name a few.

For these organizations, Tableau is emphasizing the importance of its live querying functionality, as it taps directly into sources (as opposed to querying local extracts) and therefore is completely up-to-date and auditable. The vendor is also working with extract-based technology partners to add lineage to extracts of data in the future. Tableau has expanded the remit of its native data-blending functionalities, which is positioned at all ad-hoc blends, while recommending formalized ETL/staging/normalization-type processes to use solutions from partners such as Informatica, Trifacta, and Alteryx. This is "co-opetition" at its best: Tableau doesn't want to alienate its data management partners, but at the same time, wants to equip users with out-of-the-box tools to help shape their data prior to or during analytics. None of these features is exclusively for the large enterprise; small and medium enterprises may in fact benefit more from these.

Given that large enterprises rarely standardize on one tool, Tableau is open to others using its metadata – which is in XML, so customers can load it in any third-party metadata management system and use it. In addition, there were a few significant announcements around upcoming enhancements in data management technologies, but unfortunately these are under NDA. Expect to see more commentary from us when these are market-ready. Overall, I came away with confidence that I could safely recommend Tableau to our enterprise customers, even ones with complex analytics needs and a diverse technology stack.

What next?

Tableau, like many of its competitors, is trying to develop a platform that helps harmonize data across multiple sources and deploy it anywhere (on-premise, cloud, hybrid), while being performant, and at scale. The intent is to create a logical data layer and a simplified deployment environment, while allowing IT to secure and govern it as non-intrusively as possible. This is an ambitious goal and to be fair, not even mega-vendors have a winning formula sorted out today. However, Tableau seems to be further than where most of its competitors are today, and resolved to see it through.