What Happens in Vegas is TDWI and the Future of BI


Las Vegas Sign photo courtesy Esquenta.com.br.

In late February, I had the pleasure of attending TDWI Las Vegas as an observer. A key differentiator between TDWI (The Data Warehousing Institute) and many of the other tradeshows I attend in the Big Data and Analytics space is that TDWI focuses on deep subject matter expertise taught by luminaries such as:


The quality of education leads both to a more informed end user audience, and to more interesting conversations in class, in the media room, and on the presenters’ floor compared to other shows that I attend.

In that context, I found it interesting to compare Jill Dyche’s keynote, “The New IT: How the Next Wave is Changing BI” with what I actually saw from presenters and sponsors to see if vendors are getting the message and how end users need to close the gaps.

Although you should watch all of Dyche’s great talk, there are several points that especially resonated with me.

First, although businesses have the challenge of changing with their employees and transforming their technology and infrastructure faster than ever before, there is a counterchallenge from IT, financial, and HR executives who may regard new technology deployments as little more than an excuse for resume building. New technologies can be viewed as an attempt by employees to simply build their own skills for their next job, rather than as a capability that will improve corporate outcomes.

This mindset towards technology is indicative of the traditional world of IT where you can “set it and forget it,” and technology assets are seen as permanent investments. However, this “old world of IT” attitude is less relevant to business than it ever has been before. As technological evolution accelerates and corporate pressures increase, companies must stay nimble and be willing to try new solutions that will scale and include the insights that companies need on a forward-facing basis. Dyche recommends including new data or functionality every 3 to 4 months in your BI environment as part of this process.

Second, BI maturity currently has less to do with technology and everything to do with delivery. DataHive believes that there is a new era of business intelligence arriving where network analysis, data topology, contextual sentiment analysis, facial and video analysis, metadata creation and curation, cognitive intelligence, dynamic visualization, sensory inputs, and advanced data engagement will augment traditional quantitative and statistical analysis. However, we are still a few years away from an environment where all of these tools are readily available since these are admittedly future-facing capabilities. (As a hypothetical example, consider that humans can smell one part per billion of certain compounds. This would be considered a “Big Data” detection issue for traditional visual and analytic tools, but we take it for granted as a sensory ability.) Current analytic environments are still best served by providing the analytic tools that have existed over the past 30 years to a broader end user audience that is seeking greater insight from data.

Third, business is getting its own IT funding and Millennials have less patience than previous generations. Both of these trends demonstrate the desire for departments and individuals to take control of their own IT, rather than wait for direction and guidance from a centralized IT function. These trends lead to a business environment that seeks faster delivery for personalized and specific functionality. For business end users, it is much more important to support specific needs than to implement a platform that supports holistic enterprise needs. Although we may solve this challenge with a platform, we also need to identify and enable the point capabilities that will end up supporting our key users.

Because of this change in technology, Dyche pointed out, “who thinks IT is strategic anymore?” CIOs want to shed commodity technologies that they are known for supporting. Strategically and selfishly, we all know that once our skills and responsibilities become commoditized, we are no longer critical to the success of our company. Anyone who has ever been outsourced or laid off can understand this pressure very personally; CIOs are no different from anyone else in trying to avoid this trend. In today’s environment, strategy and culture drive CIO decisions more than traditional technology. This means that CIOs have to be service brokers and service integrators simply to stay current. The broker model is no longer a strategic differentiator; it is now table stakes for CIOs. Technology is ubiquitous. It makes no sense to fully centralize IT.

Fourth, IT-business alignment is more important than ever, but it is also more of a cliche than ever. Every IT department claiming that they do it, but many are just providing lip service without putting appropriate business criteria in place to assure alignment. In Dyche’s key note, she spoke about a regional bank CIO that made this alignment concrete by defining four basic criteria for new projects: decrease cost structure, increase customer satisfaction, increase customer acquisition, or increase revenue per transaction. Any project outside of those goals would not get funded. This is the other side of new IT: in addition to being personalized, it also has to affect top-line or bottom-line growth. We no longer live in a world where technical excellence or academic curiosity can drive big technology projects. There has to be a business initiative that drives analytics and technical projects. You can be a profit center or a cost center: choose to be a profit center.

So, given these key trends, how did the vendors compare to both this vision and presentation of the future of IT and analytics? I’ll tell you who stood out to me both based on product and end user reactions later this week.

2 thoughts on “What Happens in Vegas is TDWI and the Future of BI

  1. Hyoun,

    Hi Hyuon,

    Awesome article. I think there are a lot of interesting changes happening in the analytics space (not necessarily just in advanced analysis), but particularly in terms of architectural advantages, web-based approaches, and rapid iterations on analysis that gets closer and closer to realtime reporting.

    I’m curious if you had a chance to stop by the Looker booth in TDWI. It’s a big departure from what I was doing at Ayasdi, but in the BI space it’s seems to me to be an inevitable replacement to old fashioned BI tools.

    Briefly we’ve freed people from ETL/cubing/datamart world, by running totally in-database. Our technology wouldn’t have been able to exist without the advent of blazing fast DBs like Redshift, Vertica, Greenplum, etc. But with high powered (and cheap) databases, we’re seizing the opportunity to make BI scale directly with the size of you backend DB. I mean, if you have invested in a SSD Redshift backend, why on earth would you want to transform that data and move it? We have a modeling layer that lets you do all those transforms at query-time, in the database — and we even circumvent the in memory portion by allowing you to write transformations to a temporary table that can persist in database and that can then be joined in across your model.

    If you didn’t have a chance, it may be worth a look. The company was founded by Lloyd Tabb who wrote the first web application server and was a heavy hitter at Netscape back in the day.

    1. Jeff,

      Thanks! I’m finishing up a draft of the vendors that got my attention at TDWI. I had a good chat with Keenan and am just running behind due to a couple of consultancy responsibilities. (Clients come first!) I look forward to sharing my thoughts there, but my quick opinion is that in-database is the next stage of getting analytics closer to the data. (Or actually in the data in this case!) It’s especially interesting to us because of our focus on Social Big Data where the collection of time-series data is better served with in-database analytics rather than simply performing in-memory.

      Hyoun

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>