There, I’ve said it. I think it’s a question that needs to be asked because the answer isn’t a clear one to most of us. Data discovery is a rapidly growing part of the business intelligence (BI) market. In fact, millions of dollars are being spent on these tools annually. But the question remains, is data discovery actually a good thing?
I continue to be perplexed by the rise in popularity of data discovery so perhaps I should take a moment to explain my thinking.

George Satanyana famously said, “Those who cannot remember the past are condemned to repeat it.” So, let’s take a walk through (recent) BI history.

In the early days of the BI market, BI was typically deployed on a departmental basis. Early projects were most often based in the finance department. As time went on, though, the use of BI increasingly spread to other areas of companies where, much like the finance department, data was vital to understanding performance. Some of the most popular areas from the get go included sales and marketing.

At that point in time, companies were talking about the need for organizations to manage information in a more holistic way because often these tactical deployments created silos of information based around the core departmental application, which ultimately prevented valuable cross-departmental visibility. While every department would have its own set of measurements, executives didn’t have access to consolidated and consistent metrics across their entire business. The secret to the success of any BI implementation is that key — and often elusive — single version of the truth. For IT, the proliferation of systems, tools and vendors meant that BI architecture became complex and expensive to manage.

This led to two developments: the architectural idea of the BI platform where a range of different BI applications from reporting to dashboards could be hosted on a single system performing all of the administrative and server functions and the concept of BI Standardisation with an associated drive to delivering a simpler BI architecture, which in turn had a positive effect on ease of use, training, usage, consistency and Total Cost of Ownership.

It’s important to point out that at Birst, a BI platform doesn’t dictate a one-size-fits-all approach – we cater to a range of users in terms of technical ability and role by providing a range of BI interfaces for different needs. This runs from pixel-perfect reporting via ad-hoc reporting to dashboards and yes, data discovery! Naturally, the option to consume information via a mobile application is also supported. The key, of course, is that these are all beautifully integrated because Birst was designed as a web application from the start.

You can probably see where I’m going here because the rise of data discovery seems to be recreating, at best, a departmental approach to BI, and at worst, we are creating information silos of one where each analyst in the organisation creates their personal view of the business with no governance, no rules, and ultimately information anarchy.

Are we treating the cause or the symptoms?

Before I’m accused of acting like British Mediaeval King Canute who (in)famously tried to hold back the tide and inevitably failed (yes I know that’s a popular misinterpretation, but bear with me), we should look at some of the reasons behind the new-found popularity of data discovery. A few years ago, I thought the global recession was a factor. IT budgets were shrinking and you could sense that with business survival the priority then strategic, joined-up BI investments were likely to be perceived as a luxury.

It’s now clear that the popularity of data discovery has largely been driven by the business becoming frustrated with the lack of agility of IT in delivering BI solutions. Because these were typically delivered with legacy BI systems, users would often wait months for their requirements to be met. Rather than do that, they have been declaring independence and solving the problem themselves. This is a classic case of treating the symptoms rather than the cause. What if we could address the agility issue at the cause rather than building hundreds of analytical silos?

The answer is in the cloud

Today, organisations need agility more than ever and therefore should be looking at cloud based analytics from Birst. “Cloud BI”, delivered on a Software as a Service (SaaS) basis, can enable departments to get up and running very quickly, without waiting for IT to install, configure and endlessly patch ancient legacy BI systems. Inherently scalable, the number of users can be easily increased to accommodate a growing business and functionality can grow as user requirements increase. Something data discovery cannot do. Birst’s automatic data warehouse generation reduces at a stroke the most time consuming part of any BI project.

At this point I should point out that Birst has, of course, launched a data discovery tool called Birst Visualizer. What do you think about that, I hear you say? Well, I’m all for powerful data discovery tools as long as they provide governed data discovery with access to a consistent and managed data source and are seamlessly integrated into a BI platform that provides everything from reporting through to dashboards. That way we can meet the needs of users to powerfully explore data without creating anarchy. I would also point out that Birst Visualizer was designed from the ground up for business users, not data (rocket) scientists although rocket scientists are welcome to use Birst, too!

A BI platform like Birst means that organisations will still enjoy the benefits of a unified view of their business and can switch on functionality as and when it’s needed. Recent Gartner research shows that 49% of respondents have either deployed Cloud BI or plan to do so in the next 12 months and they estimate that by 2016, 25% of all new analytics projects will be cloud-based, so it’s clear that many have already discovered that this is the agile answer to legacy BI.

Instead of throwing away all of the painful lessons of the past and implementing information silos to solve your BI agility issues, try a born in cloud BI platform like Birst. It’s fast to implement, easy to use and unlike those data discovery tools, really does scale with your requirements.