Business intelligence has historically failed to deliver, whilst BI vendors have constantly sought trendy new bandwagons to jump on. Does this mean that companies should abandon the hope that software can provide business intelligence? Fortunately, the future looks bright argues Andy Honess.

By Andy Honess, QlikTech
When something is so obviously essential such as food, clothing or shelter, that thing will naturally create a market. If Abraham Maslow had constructed a hierarchy of needs for businesses, ‘intelligence’ would surely have been at the base as a fundamental need for survival. The alternative is either business stupidity or business disinformation, depending on your grammatical interpretation, but neither a desirable state of affairs.
The term ‘business intelligence’ also just sounds like something that only heretics would disavow. It implies a kind of virtue like other trendy corporate buzzphrases like ‘employee engagement’ and ‘corporate social responsibility’.
Thanks in part to this heady cocktail of essence and implied virtue, the BI club has grown over the last 20 years to include as its members all manner of enterprise software snake oil merchants and every year new vendors jump on the bandwagon. Who wouldn’t claim to provide ‘intelligence’? This takes me back to the height of the dotcom boom when so many vendors were slapping an 'e' before their company names and sorting out the details later.
Another practical reason vendors have warmed to BI is that it has topped the league tables year on year as a top CXO priority. The most influential of these is the Gartner EXP Worldwide Survey of 1,500 CIOs, which once again put BI at the top of the table in their most recent survey last January.
In an effort to differentiate themselves, BI vendors constantly seek trendy new bandwagons to jump on. When high-profile corporate scandals broke around the world, vendors vigorously starting touting compliance. When the credit crunch rocked the financial markets, BI promised to deliver transparency and accountability across organisational silos. These potential applications of BI were hyped up regardless of whether the tools could realistically deliver by greying the distinction between basic product functionality and potential business benefits. This is about as ingenuous as promoting a biro as a tool for producing a best-selling novel. True, but only in theory.
To attract the younger generation, there are new bandwagons to jump on in the form of ‘mash-ups’ and visualisation which are all about bringing data together and presenting it graphically. The kids love it and customers demand it so we better tick those boxes as well!
For many customers the hype and ambiguity surrounding BI has ultimately led to more despair than genuine intelligence. In reality, traditional BI products aren’t capable of practically solving these complex and highly situational problems in timeframes that are meaningful to businesses. In fact, BI has historically failed to deliver even basic information like key performance indicators and business process management. The few that did derive real benefits from these tools tended to have to be in the upper echelons of leadership with access to ample IT and budget resources.
Self-service BI environments
When the BI space began to emerge 20 years ago, memory was expensive and processing speeds were slow. As such, entrants into the space were constrained in how they might technically approach so-called multi-dimensional analysis. Specifically, they were forced to employ a 'pre-calculation' approach that involved the use of 'cubes' to organise and store the results of these pre-calculations. While the results appeared to the user to be instantaneous, the main challenge with this so-called 'online analytical processing' (OLAP) approach was around the time and cost associated with defining and redefining cubes. All this resulted in a process that was too complicated and time consuming to be practical.
Does this mean that companies should abandon the hope that software can provide business intelligence? In fact, quite the opposite. Today there are more software vendors today that stand to deliver on this promise than ever before thanks to massive leaps forward in processing power and usability.
Gartner defined a new breed of ‘self-service’ BI vendors aimed at the non-technical business user that are characterised by low pricing models, very short implementation times (hours or days), minimal training requirements and in-memory processing. In short, fast and simple. In its report ‘Emerging Technologies Will Drive Self-Service Business Intelligence' Gartner even went so far as to warn IT departments to embrace these new upstarts and integrate them into their IT strategies or risk being forced into submission at a later stage.
The report predicts that by 2012, emerging technologies such as interactive visualisation, in-memory analytics, search, SaaS and SOA/Web 2.0 will marginalise IT's role in building BI applications. You could look at this from another perspective and see these trends as empowering the end users.
In these ‘self-service’ BI environments the act of deriving knowledge from data becomes literally contagious because users get instant gratification and tell their peers. At the risk of jumping on a trendy bandwagon, we have already started to see people using social networking tools like Twitter and Facebook to share information on the many ways that QlikView can be used from traditional applications like KPIs to unconventional things like looking at police crime statistics.
Speed and simplicity have created the impetus for a genuine revolution in people’s ability to view and analyse massive data sets. Speed and simplicity have been at the heart of all major technology revolutions from the automobile to the Internet. The reason that the internet is transforming the publishing landscape is because it has removed obstacles of technology, cost and speed enabling the masses to publish, duplicate and disseminate information for free. This democratisation of the formerly exclusive and expensive task of publishing has forced new consumer-friendly legislation, threatened dictatorial regimes and broken news stories that might have otherwise been covered up.
In a similar vein, the new self-service BI vendors provide a low cost, fast and easy option for people who want to view, analyse and share corporate data. This ultimately creates a strong competitive advantage in a world where it is becoming increasingly difficult to eke out further profit margins through old standbys like restructuring and cheap foreign labour.
From this perspective, business intelligence is only the tip of the iceberg. As more and more people gain the ability to view, analyse and draw their own conclusions from massive data sets, this has the potential to stimulate extraordinary progress in areas like media, medicine, scientific research and government policy. Everyone’s invited to jump on the bandwagon!
Andy Honess is UK MD at QlikTech.
MyCustomer.com 09-Jun-2008
Story read 3051 times
I think that not only will 'BI in the cloud' assist any users in DIY BI, but there's also the fact that you can buy commodity data-centric super computers now that will fit under your desk! This enables incomprehensible amounts of data to be processed faster than ever before, but by the 'power user' themselves - i.e. without the help/hindrance of IT...
What do you think this will do to consistency of information when every man and his dog can do their own BI...? I think it's an emerging challenge for every corporation to consider.
The delicate balancing act of enabling the freedom of DIY BI, but also maintaining the 'corporate/sanitised' version will certainly not be made any easier by this trend, but it is essential to innovation, that's for sure.
Phil
Information Consistency
Of course there is plenty of debate that the flood of what I call 'Excel Anachy' and the 'multiple versions of the truth' could blight new generations of self service BI tools and lead again to high risk decisions being made around the misuse of Excel (personal errors or inputs). So, of course, controls are as important as ever when you are delivering business critical applications/information to an organisation, but QlikView is a great example of the ability to validate the truth, validate the source of the data - by the specific capability to drill down to the individual transaction or record that is the basis of your assessment or exploration or conclusion. (All data, to the lowest level is held in high volumes in memory (RAM) allowing the user to explore and validate whereever and however they may be provoked).
I strongly believe this technical capability (instant drill down) is the foundation of TRUST which is so vital to (1) the success and uptake of a BI tool and (2) the cultural change required to make an organisation a predictive and lower risk decision making entity.
So I do believe that you can liberate the individual user to explore and become 'empowered' with company and personal data but as long as you have the necessary controls on the corporate data and the abililty to drill and validate what you are looking at, then the two worlds can exist in a very efficient, very exciting new generation model for BI and analysis in general.
Andy.