Update on Top 6 BI Trends for 2012

Analyst Reports, Big Data, Big Data Analytics, Cloud BI No Comments

Late in 2011, Gartner came out with their top 6 predictions for business intelligence: BI in the Cloud, Mobile BI, Analytics, Agile BI, In-memory analytics, and Big Data. As we enter the final quarter, let’s take a look at how things are panning out.

BI in the Cloud

As Cloud computing continues to dominate the IT landscape, Gartner is predicting that cloud offerings will make up just 3 per cent of BI revenue by 2013. However, it is gaining ground and chipping away at on-premises BI, particularly in the Big Data Analytics space.
Decision makers are still questioning the Cloud as an operational tool. The initial move of data to the Cloud is a challenge, as is bandwidth and security. Many are favoring using Cloud of back up storage and archiving, and disaster recovery, but are still unsure of how well Cloud will support operational data activities.

Mobile BI

Forrester’s believe Mobile BI will go mainstream in 2012. Whilst I agree that mobile BI will continue to grow, the small device platform is suited only to a limited subset of business intelligence activities. There is still an issue around data security on caching data on mobile devices, which are readily lost or stolen. There are also processing power constraints, limiting any real value with analytics.
As a quick, on the go monitoring tool – great – but I am unconvinced that mobile devices will suit the escalating swell in analytics needs.

Analytics

The MIT Sloan Management report found that organizations using analytics are more than twice as likely to substantially outperform their competitive peers. Analytics is certainly the one to watch. As many companies are still struggling to publish dashboards on historical data – the technical needs around advanced analytics are still well beyond many organisations. This is certainly one for the Cloud – and watch out for a rapid explosion of Analytics as a Service offerings. Moving analytics closer to the data is a key strategy, and with Big Data better suited to the Cloud – it makes sense that analytics will follow.

Analytics has already taken off, it is now a matter of maturity towards optimization and prediction becoming integrated into business processes so that it is consumable as a decision management tool.

Agile BI

Agile is good with anything today. Methodologies are not getting a lot of focus as BI teams are getting swamped with requests for reports and dashboards that are better off being done locally, but business users. However, sadly too little attention has been given to training in using BI tools and in effective dashboard design. Before agile BI can really take hold, BI teams need to get more aligned with their real function – and pass publishing back to the business.

In-Memory Analytics

Memory continues to get cheaper by the day – and in-memory analytics tools such as Qlikview, Spofire and Tableau are helping users gain more insight into the power of simple data exploration. The real benefit, is that local data sets can be used, rather than relying on waiting for IT to import data into centralized data warehouses. This does cut out the data quality look – but for exploration purposes it’s a good start to helping businesses to expand their business intelligence portfolio.

Big Data

Forrester’s prediction that Big Data will move out of the silos and into enterprise IT may be right, but its not the best decision for many companies. With the rapid growth in the volume and variety of data it makes much more sense for Big Data to be in the Cloud. Corporates have enough to do without continually adding more servers and memory.

Big Data is certainly changing the business landscape – however, its maturity has yet to cross ‘The Chasm’ and corporate IT is better to wait until it has, and for their business users to get more savvy about how to use Big Data.

Big Data Analytics – Moving From Prediction to Possibility

Analytics, Big Data, Big Data Analytics No Comments

To date, BI analytics tools have largely been used on historical data to predict likely outcomes of various use case scenarios. In such instances, the use of historical data adds a dimension of probability to the statistical realm.

Big data is not historical data – it is Now Data – and rather than be bound by prediction, or even by probability, Big Data Analytics takes a look at the possible.

One of my gripes with so-called ‘men of science’ is that they refuse to accept that something is possible unless there is a way to define it in terms of empirical evidence that can be replicated with consistent outcomes. Whilst I understand the value of ’empirical science’ for many instances, I don’t feel comfortable with such a closed-minded approach to everything.

My belief is – ‘just because no-one has yet determined how to measure something, or indeed replicate its existence or impact, doesn’t mean to say that it doesn’t exist’, or is not possible.

Finally with Big Data Analytics, I feel my view is gaining substance. There are a number of scientists today researching Quantum Physics who also stand beside me – they know that certain things occur in fields of energy, they just cannot replicate it in a way that enables them to define it with certainty or at least control it with reasonable probability.

Are we finally moving along the continuum from Prediction > Probability > Possibility?

Keep Your Eyes on the End Game

Big Data Analytics, Cloud Computing, Dashboard Design, Data Visualization No Comments

In business intelligence it is easy to get caught up in the data technology, the current obsession for ‘big data’, or producing as many dashboards as possible. However, none of this is as important as the end game  – extracting usable insight that will improve your ability to reach your strategic goals.

It frustrates me that BI vendors continue to pump out dashboard building tools that fail to deliver the basics of good dashboard design, yet they seem to have the resources to deliver new tools for big data. Whilst I am a great supporter of the ability to be able to extract market triggers, and explore data for possibilities, rather than rely on historically driven predictive analytics, there is a danger that we are losing sight of what we are aiming to achieve.

There are those who claim that big data is nothing new – are they missing the real value that Big Data delivers, and again, focusing too much on the technical definition, rather than the marketing definition? Data volume has always been an issue – and yes, virtualisation and cloud computing have certainly improved processing speeds. In-database analytics, and in-memory processing have also helped us get to the truth faster – but how fast do we really need to go? As humans we can only view, assimilate and make decisions at a limited speed. Rather than faster data – we need smarter ways of making decisions – robust decision models, statistical models to trigger alerts, to provide us with glimpses of a likely future outcome. Big data [as defined as social data and other data plucked from the marketsphere] is being made available in ways it never has been before. That is big news! That is a big jump from the market intelligence channels we once had to rely on. The key is in finding the value within this data.

I love technology innovation – but not at the expense of sound logic, readily dispensed, and in a format that is rapidly assimilated and instantly usable.