Designing the Ultimate Business Intelligence Tool
I was recently contacted regarding a blog by Jaime Brugueras(1), discussing what he feels is missing from the current generation of Business Intelligence (BI) tools. They asked me to provide my feedback via a blog post and hopefully it would start a discussion.
Everything that Brugueras describes in his blog would do including the latest BI tool. It clearly highlights key pain points felt by all user levels and creates the framework through which these could be addressed. Despite his observations, I feel that the nature of the market and the current landscape of BI tools prevent these recommendations from being carried out.
The central argument of Brugueras’s blog can be summarized as follows: Business Intelligence tools must be easier to use, more comprehensive in nature, predictive of future results, and cheap enough that even the smallest companies can afford them. . He goes on to advocate for tools that are more sophisticated than a team of IT professionals could program, yet simple enough for a layman to set up.
He notes that “the tools available today are relatively complex and require some level of programming,” however, a few lines later he states that “an effective BI tool enables seamless integration of data across…software.” of CRM, accounting and point of sale”. However, he laments that current BI tools “cannot integrate data from all sources.”
While various tools can generate similarly formatted CSV files, the database from which these exports originate does not have the same schema. There is no universal key to link multiple data sources, and few companies are willing to become information providers by making it easy for others to use their data.
Many would dismiss my previous statement calling my attention to APIs; how companies like Google, Foursquare and Twitter make their data available to the outside world. They would be correct, except for one key point: the user accommodates, rather than dictates, the format of the API. If you don’t like how Twitter named a particular variable, that’s a shame for you. A company like Google is not going to change its data structure just because you ask nicely.
When you consider that every company has its own special flavor of APIs, the dream of “an effective BI tool (that) enables seamless data integration” simply goes up in smoke. Every time a vendor comes along, you’ll need to adapt your BI tool to accept their data or ask them to conform to your standard. The former is much more likely than the latter, but doing the former requires programmers, and programmers cost money.
It’s not hard to see the relationship between cost and compatibility. Being more compatible requires a larger developer base, which in turn requires more capital. The resulting tool would have to be heavily supported by advertisements or sold at a price sufficient to cover ongoing development costs. It quickly enters the realm of enterprise-grade solutions, where even traditionally free Google Analytics started charging $150,000/year for a Premium level of service. This price is clearly out of the affordability range for most small businesses.
The author advocates for a comprehensive and low-cost solution aimed at the SME market. He wants a tool that is easy to use, inexpensive, and easy to use, that features automatic integration with multiple data sources that predicts future outcomes. He acknowledges that end users “probably can’t program their needs,” but advocates modular software development in anticipation of all possible needs. These BI tools must be action-oriented, distilling complicated tasks like customer retention, inventory management, and social media communication with the simple click of a button. While it’s a wonderful concept, I don’t think everything is going to be contained within a single tool.
Apple products, whether it’s iPhone, iPad or iPod, are famous for working well together. Apple accomplishes this by controlling each step of the process and knowing exactly what each piece of hardware includes. They can then perfectly tailor their software to work within the hardware’s specifications, producing a very compelling product offering. However, a Mac fails when trying to run software originally written for a PC. With Apple reluctant to license its iOS to third-party developers, don’t expect to see this unit of the product line increased or replicated any time soon.
There are only two ways Brugueras’ ideal BI tool could come to fruition: a unifying open source project of unprecedented scope, or a single (for-profit) company willing to take users from cradle to grave to all your business software needs.
While I don’t think a massive open source project could ever appear, I insist on never saying never. However, the cynic in me just doesn’t see it. It’s also not very likely that a single company will create a full-featured A-Z Business Intelligence tool. The barrier to entry is not high enough to someone I wouldn’t try to come up with a “me too” product offering.
I’m not here to preach about what my ideal business intelligence tool would be, because I don’t think there can be a one-size-fits-all tool. You’ll never get enterprise-grade features in a cheap or free product, either because of the associated development costs or the simple inability of small businesses to spend the time necessary for such a powerful tool.
Take your prototypical small business as an example, where employees typically wear more than one hat. Your “web guy” is likely not only responsible for website design and SEO, but also PPC and SEM and possibly copywriting for both the online channel and traditional corporate communications. Ask yourself, will this overworked person be able to spare the time to use a highly sophisticated BI tool?
Just like clothing, creating something that is one size fits most often results in a garment that doesn’t fit well on 99% of the population. When buying a suit, most people require that the pants be hemmed or that the jacket be put away. Few people are truly “off the shelf,” so why do we expect the same from our BI tools?
There are tools out there that work great for small businesses, that scale well, and can adapt to the business as it grows. They don’t have the features of an enterprise-grade tool, but small businesses don’t have the bandwidth to use all those features anyway. Instead of spending time chasing the dream of the perfect tool, end users should focus on better expressing their BI tool needs. The business world has been quite successful in recognizing needs and designing products to meet them. I have every confidence that a number of different vendors will step up and provide specific solutions, as long as the requirements are clear and properly documented by the end users.
(1) A link to the Jaime Brugueras Blog