A little knowledge is a dangerous thing – and that’s especially true in information management. If you only know part of the story, you can easily make a bad decision that takes your company down the wrong path. It’s only by getting the whole story from your data that you can responsibly guide a company.
BI and analytics software can help uncover this story, but there are now hundreds of companies offering technologies designed to dig into data. This proliferation of tools is both a blessing and a curse. While competition breeds excellence, there are a few standards or best practices embraced across the board. As a result, the onus is on individual businesses to embrace and uphold policies that will enable the effective use of data in a responsible, governable way.
One increasingly attractive solution for doing data right is to leverage a visual analytics platform. Unlike standalone data visualization tools (which can provide useful but sometimes misleading views of the enterprise), a visual analytics platform weaves together all the elements of a full technology stack. This enables analysts to see the big picture, but also to drill down and across the data landscape, to not only understand what’s happening – but also why it’s happening.
What defines a “platform”?
The term "platform" gets thrown around a lot in tech – what does a platform really mean, in the context of visual analytics?
1. A platform acquires data.
Data integration has long been a challenge in BI and analytics. It was true back when companies were only dealing with a handful of important systems, and now, the explosion of cloud-based systems has made things exponentially more challenging. In marketing alone, there are more than 3,500 cloud-based solutions on the market. And while there are similarities between these applications, the reality is that they all generate different forms and formats of data.
Traditionally, organizations used ETL software to load data into data warehouses to deliver a single, unified view of data. This led to tremendous duplication of effort, which resulted in unnecessary costs, inefficiencies, and most seriously, data quality issues.
A first-class visual analytics platform will therefore include built-in data integration to help improve quality and lower cost. By marshaling data acquisition through a single platform, every aspect of the process becomes both simpler and more governable. Data governance in particular has become a major touchpoint these days; but without a centralized, manageable platform: governance is all but impossible.
2. A platform allows for freeform discovery.
Freeform discovery is critical to data analysis. The key requirement here is to enable a "conversation with the data". Here, business users can start with one query, get a good visual representation of what the data reveals, then quickly iterate to get to an increasingly accurate picture of the data that really gives them the answers they’re looking for.
As the business world has become more multidimensional, so have the visualizations to understand that business. Analysts need the ability to add dimensions, remove them, and re-render them to get a better view. This "data conversation" can sometimes take 20 or more refinements to really see what's happening. If the user has to wait minutes or (gasp) even hours to do that fine-tuning: they’re less likely to do so, and things go undiscovered.
As a result, a platform with built-in indexing capabilities is really what’s needed to address the requirements of curious users. In the past (and even today) sampling has been a common tactic for doing analysis without needing to crunch all the data. But experienced analysts know that sampling is a dangerous game. The wrong segment of data can be misleading, causing inaccurate assessments, and bad advice.
3. A platform offers a robust visual library.
Time was, business users trying to demonstrate their vision would rely on the three horsemen of data visualization: pie charts, bar charts, and line charts. Obviously, the world doesn't fit into three simple visual categories any more than square pegs fit into round holes. Fortunately, as analysts and software vendors have matured we've seen an explosion in the types of data visualizations that can be used to understand an increasingly complex world.
A solid visual analytics platform should therefore come with a built-in library of data visualization types, upwards of 40 or more, which can be applied to any number of datasets. Arc diagrams to represent networks. Area graphs to show the development of values over time. Flow charts to track the movement of customers, products, or markets. In short? Having the whole story is one thing – you still need the right tools to tell it too.
4. A platform is extendable and embeddable.
Oh, the days when every piece of software was installed on-premise! Sure, life was tough for the IT team, staying up-to-date on all the latest releases. But at least you knew where your stuff was. Today's analytic applications must extend far beyond the corporate firewall. End-users are everywhere with their mobile devices, ready to dive into data.
Where ETL and change-data-capture were once the bread and butter of data-driven applications, now there's a whole new world of opportunity to incorporate data from cloud-based systems, or third-party data accessible via the cloud. And to access this type of data, APIs rule the day. A robust data visualization platform must natively embrace APIs and API management.
This isn’t the only way a platform must extend, though – no, it also needs to keep up with the changing needs of your business, scaling with you as you grow. Growing businesses can take advantage of cloud-based platforms like Qlik Sense Cloud Business, as they allow organizations to quickly and easily update capabilities as requirements change – without having to perform major installs or updates.
By taking advantage of the expanding, multi-faceted world of today's information worker, a visual analytics platform can reach much further into the various domains of business people, prospects, and customers, creating a whole new realm of possibilities by enabling analysis anywhere, any time, with any kind of data.
Want to see what a true visual analytics platform looks like in action? Check out this on-demand webinar (hosted by me!) and featuring Qlik’s Solutions Expert, Daniel Brault and Bloor Group Chief Analyst, Dr. Robin Bloor.