On May 13th, Forrester analysts Boris Evelson, Jim Kobielus, Gene Leganza, Holger Kisker and Noel Yuhanna joined me in hosting a data management TweetJam on the topic “What BI is Not!” using the hashtag #dmjam. (You can still see the results and ongoing conversation if you search the hashtag.)

During this one-hour TweetJam, we asked the following questions, leaving 10 minutes of Tweet-time between each question:

  • Do you prefer the broad or the narrow definition of BI? Should ETL, DQ, DW, MDM be considered part of BI?
  • How should we differentiate BI and analytics?
  • What’s the difference between business intelligence and other forms of “intelligence” like competitive intelligence, market intelligence?
  • Is convergence of structured and unstructured information hype or reality?
  • Is BI looking only through the rear-view mirror, or should historical and predictive BI be one and the same?
  • How will social media impact traditional BI?

The response to this event was extraordinary, and we have a large community of data management and BI thought leaders who joined the conversation to thank. During that single hour there were over 360 Tweets with 65 unique Tweeters actively joining the conversation (not including those who only listened). If you include Tweets leading up to the event and the continued conversation after the event, we’ve seen over 480 Tweets and over 100 Tweeters … and growing.  

But what did we accomplish (aside from providing an entertaining distraction for a number of people)? Below, I’ve summarized a sampling of the takeaways that were shared by some of our participants on each question:

1. Do you prefer the broad or the narrow definition of BI? Should ETL, DQ, DW, MDM be considered part of BI?

There were as many opinions as there were Tweets on this subject, but many agreed that while BI is highly dependent on these other data management technologies, it is not the only initiative/application/competency served. For example, ETL, data quality and MDM capabilities increasingly serve real-time, transactional processes like order management, call center operations, and supply chain management in addition to traditional BI and decision support.

Here are some great Tweet quotes shared by some of the participants:

  • TaliskerCats: @rbkarel  They r the underpinnings of BI — if you have no / bad data all the decisions you make using BI based on it is invalid  #dmjam
  • Lorita: #dmjam better to stop talking about #BI as if it handles everything and start talking about broader Information Management for all of it
  • SethGrimes: BI includes processes & tools for identifying, accessing & preparing data for analysis, working backward from desired outcomes. #dmjam
  • Santaferraro: #dmjam If we want to talk to the business, then it makes sense to put it all under #businessintelligence Make it make sense to them!
  • Chempfield: @rbkarel yes, but fear that BI as a space would become too big and unruly. In some respects it already is #dmjam
  • Cmatignon: @bevelson some do but I disagree. BI is part of the larger DM picture but DM includes a lot more: rules, processes, predictions… #dmjam

 

2. How should we differentiate BI and analytics?

The consensus on this conversation was that it shouldn’t really matter what labels we use, it’s more important to focus on what value these solutions are meant to deliver. That said, plenty of passion was shared on which labels make the most sense, with no easy answer. Some felt that the whole debate was created by BI/analytics vendors as a way to differentiate themselves from one another, while others felt it was important to frame a distinction as a way of educating business partners. In the end, I agree that the only reason to spin your wheels on clarifying this definition is if you need to clear up any confusion your business partners may have in supporting and evangelizing the business case to invest in these capabilities.

 Some great Tweet quotes shared by some of the participants:

  • Gilliatt: I got the impression that #analytics is #BI, shorn of its legacy baggage from existing products/deployments. #dmjam
  • Actuate: @rbkarel BI is the act of data aggregation and delivery to end users. Analytics adds math, forecasting- the life to the user question #dmjam
  • PerformMagazine: There is no difference – a way to sell more #analytics and #BI sofware?? #dmjam
  • Actuatetweeter: #dmjam  #BI covers the entire spectrum of BI skillsets – thus nobrainer that anlaytics is a subset of BI.
  • NeilRaden: #dmjam Is Buffy the Vampire Slayer comedy or gothic romance? Does it matter?

 

3. What’s the difference between business intelligence and other forms of “intelligence” like competitive intelligence, market intelligence?

Relative to some of the other questions, this topic had the most consensus overall, but that doesn’t mean everyone was singing Kumbaya. That said, the majority felt these other forms of intelligence were just functional- contextual-oriented subsets of broader BI solutions.

 Some great Tweet quotes shared by some of the participants:

  • SethGrimes: Competitive intelligence, market intelligence, etc. are particular functional applications of BI #dmjam
  • Gilliatt: BI is an application of technology; competitive and market intelligence are business functions. Apple and oranges. #dmjam
  • Birster: #dmjam What's the diff btwn #BI and other 'intelligence' e.g., market intelligence; Apps built on top of BI for particular business needs.
  • Endeca: depends who is doing the intelligence — computers, people, or both in cooperation, or Human-computer Information Retrieval (HCIR) #dmjam

 

4. Is convergence of structured and unstructured information hype or reality?

I was personally extremely interested in the takeaways from this topic because I hear so much hype from the vendor community about how important it is to manage and provide access to all forms of information – structured data, semi-structured data, and unstructured content – in a seamless, integrated user experience. But when I speak to my Forrester clients about this convergence, they are philosophically interested but their realities dictate they need to improve the management of their structured data or content siloes first before they can even consider convergence.

 Some great Tweet quotes shared by some of the participants:

  • EventCloudPro: #dmjam I spend most of my time figuring out how to process related structured & unstructured info – extracting structure? – using #cep
  • SethGrimes: @rbkarel Analysis of info from "unstructured" sources, independently & in conjunction with DB-sourced data, certainly is reality #dmjam
  • CMatignon: @rbkarel Reality! Social media is producing unstructured data that needs to be taken into consideration #dmjam – you meant need or solution?
  • ashish_bhagwat: Unstructured-ness creeping into everything – #BPM, #CRM, #BI… Signs of the times to come – Unstructured or mash-up Enterprises? #dmjam
  • Endeca: unstructured is full of latent structure — entities, phrases, grammatical structures. extract and refine into structure #dmjam
  • CurtMonash: #dmjam Tabular and text data have radically different levels of precision (& in some cases recall). Tough to integrate tightly.
  • IBMCognosCTO:  #dmjam The most interesting part about analyzing text is having to deal with the uncertainty of interpretation.

 

5. Is BI looking only through the rear-view mirror, or should historical and predictive BI be one and the same?

The prevailing wisdom on this topic is that BI does in fact have a responsibility to deliver insights to the past, the present, and the future – with the recognition that that is often easier said than done. While predictive analytics is an extremely hot topic now, across both vendors and end users, the reality is most organizations cut their BI teeth on historical reporting before they take on the more complex challenge of modeling future behavior and trends.

 Some great Tweet quotes shared by some of the participants:

  • Birster: #dmjam  Should historical and predictive BI be same? Both are BI, but few co's do predictive yet.  Co's typically start w/historic view.
  • IBMCognosCTO: #dmjam The entire timeline of information is important: past, present and future. The battle is to be effective across it all.
  • Dougmow: #dmjam if you ignore part of the timeline is it only half intelligent?  BI/2
  • ReduxOnline: @dougmow yeah, but you should be able to rewind, tweak then fast forward again for predictive BI #dmjam
  • NeilRaden: #dmjam Folks, BI was invented for Finance and Mktg to understand results. Everything else is add-on
  • Karuana: If I try to build all forms (past, present, future) for my org I will fail. Must pick one; do well; iterate and build skills #dmjam

 

6. How will social media impact traditional BI?

As expected, our “finale” question triggered a lot of passion and excitement. (Prior to the event, my Forrester colleagues and I shared some of our thoughts on this topic in a blog post, “Social Media Will Play A Big Part In BI’s Future”).    The consensus was that the huge volume of data being captured in social media channels must be taken into account in your BI strategies – the challenges are when and how.  Many felt that both the technologies available and the BI maturity levels of many organizations were not prepared to effectively deal with this new source of insight, but progress is being made.

 Some great Tweet quotes shared by some of the participants:

  • Dakoller: @rbkarel #dmjam but proper tool integration is needed to make add. info. consumption easy and fast (for non-IT-people too)
  • Psonderegger: social media impact: more "unstructured" content, collaboration on insights discovered, analysis on social networks themselves #dmjam
  • Jrep: #dmjam: Social impact on BI? Someone else has my data!
  • Gilliatt: Social media drives demand for analytics outside traditionally IT-heavy roles. IT has to catch up. #dmjam
  • CMatignon: Socialytics will also become more prevalent, see how Google Analytics and social sites to profile traffic are becoming mass market #dmjam
  • Casmrv: To really take into account the social aspects, need to become better at dirty data. No more 6 months data cleaning projects.. Ready? #dmjam

 

As you can see from the sampling of great comments shared above, this TweetJam was a great way to discuss some of these top of mind topics for many BI and data management professionals. Be on the lookout for additional TweetJam events from Forrester Research across a wide variety of technology topics. And keep that #dmjam hashtag as a saved search – Forrester’s data management analysts will be hosting future TweetJams using that tag on topics ranging from Data Governance, Master Data Management, Data Warehousing, Data architecture, BI, analytics, and more!