About five years ago, when I implemented my first analytic application, life was simpler. There were a total of eighteen options for embedding analytics into a SaaS application. If you expanded the universe to include analytic platforms that were appropriate for use inside an enterprise rather than for white-labeled use with a customer, you could add a few more options, but not many.

Today the universe of analytic platforms is closer to a hundred viable options, and many of the functions that stood out as “cutting edge” capabilities have become commoditized. The explosion of choices for analytic leaders got me thinking: how would I make my choice today if I had to deliver data-driven insights to an organization? Well, I’d start with what constitutes “modern” analytics today.

Wow. That’s a lot of options for your analytics.

Wow. That’s a lot of options for your analytics.

Back then, when I made my selection of a platform, “modern” wasn’t a concern. The platform either could be embedded or it couldn’t. It was either designed for use by employees, or it was for experts like data scientists. It was either configurable to match (roughly) the theme of our company or it wasn’t. That was the extent of the discussion about the “modernness” of an analytic platform. It helped with the first cut of vendor candidates, but not much more than that.

Today, the world has changed. In that myriad of analytic platform options, there is a wide range of technologies, from those developed dozens of years ago to those that are only a year or two old. As a result, the capabilities of platforms are much more wide-ranging than they were five years ago. If I were making my decision these days, I’d start by identifying those platforms designed for the future, that have modern functionality, that take advantage of new technologies and approaches to delivering analytics. But what makes a platform “modern”? Here’s what I think of when I separate the analytic platforms of today from those of yesteryear.

Business User-Friendly

Early analytic platforms assumed that the end user would be an analyst or a data expert of some sort and this was often a fair assumption. But as analytics have become commonplace, the new expectation is that these analytics are easy to use for the average business person. You shouldn’t have to learn SQL or attend a training course to derive value from the analytics. They should be simple but powerful, allowing the user to explore the data without forcing them to learn a new mode of thinking.

Beware of analytic platforms that force the user to use code, code snippets, unique proprietary analytic “languages,” etc. to perform analyses. Look for analytics that are drag-and-drop easy, that use filters, options, and easily understood configuration “wizards.” Modern analytic platforms make it easy to use — and configure — the analytics the user will need to perform her job.

No Extra Moving and Storing of Data

A big hint that a platform is modern in its approach to delivering analytics is the movement — or lack of movement — of data. Back in the early days of analytics, databases were designed less for analytic reporting, and more for the insertion and deletion of records like required in a transactional system for point of sale operations. As the need for analytic capabilities (for which these transactional databases weren’t optimized) arose, people began looking for a better way to deliver fast, reliable analytics. The solution? Move all that transactional data, or maybe just a subset, into a secondary store designed specifically for high-performance reporting.

The approach worked well, but it came with a host of drawbacks such as the cost of storing data twice, the possibility of having the transactional database and the analytic data store out of sync, and a delay in the availability of data for analysis purposes.

To be clear, some movement of data will probably still occur. It’s likely that you’ll want to bring your data into a general data warehouse for purposes of consolidation and transformation. That’s not the “movement of data” to which I’m referring. What I’m talking about is a second move of the data. Some analytic platforms will take data, even if it’s already consolidated into a data warehouse, and move it again into their own “analytic” data store. Why would you do this? If the data has already been moved, already been consolidated, is already sitting there available for use, why would you move it yet again? The reason is simple: it’s the legacy of a platform designed before the availability of powerful data warehouses. The systems that move the data did so out of necessity — it wasn’t optimal to act on data inside transactional systems, and analytic-ready data warehouses didn’t yet exist. So the analytic platform vendors created a solution… and stuck with it even as times and technologies changed.

Now, these systems that solved a real problem that existed years ago have introduced new issues such as data desynchronization, extra cost, extra complexity, and the inability to deal with rapidly refreshed, ever-changing data. And, since data volumes are growing every day, these older systems often move only an aggregated subset of your data rather than the whole dataset. There goes your ability to see the raw data and trace everything back to the original numbers.

Today, this second data move/store isn't necessary to achieve high-performance analytics — it's a sign of a platform that hasn't kept up with technological innovations. When you’re making your choice of analytic platforms, look for a system that either connects directly to a data warehouse or connects directly to the data sources themselves to provide analytics. Skip the platforms that force a second, unnecessary movement of your data.

Self-implementable and Self-Maintanable

Let’s get something out of the way: I’m not talking about the mythical “self-service” analytics that have become the rage these days. Self-service analytics offer the promise that anyone should be able to set up the analytics and start analyzing — no help required. No, what I mean is that your company’s technical people should be able to implement the analytics stack without the necessity of purchasing expensive professional services from the vendor.

Too often, analytic vendors use professional services teams to gloss over significant gaps or complexities within their product. “You can implement our platform yourself,” they exclaim, at the same time delivering a proposal that includes hundreds of hours of services work. “Using our teams will get you up and running faster” is often code for “there’s no way you’ll be able to find your way through all of the gaps, eccentricities, and workarounds necessary to get this thing working. Better leave it to us.” For every change, every update, you’re beholden to the vendor and their services team.

It’s reasonable to expect services hours to be needed to help with edge cases and to explain best practices for a platform, but unreasonable to require help for every aspect of implementation. Look for a vendor that provides a solution that your team can implement and manage over time — without having to engage the vendor for every little change. It’s more cost-effective, faster than having to hire outside help, and ultimately, is a sign that the platform has been engineered with an eye toward long-term use by the buyer.

Flexible in the Delivery of Analytics

In the not too distant past, analytics were standalone, desktop applications that you fired up when you needed to perform analysis. If the analytics could be embedded into another app (and that was a big “if”), there was one method for the embed — iFrames. Luckily we’ve moved on from those days… or at least, most of us have moved on. While most analytic platforms have moved into the modern age, a few haven’t and still require desktop tools for the authoring of analytics and dashboard and the use of iFrames for embedding. Don’t accept this — it doesn’t have to be this way. Modern analytics are web-based for end users to use the charts and dashboards and for authors to create the analytics for the end users. There’s nothing to download, patch, manage, or update. It’s all delivered as a SaaS app making life easier on both the end users and the team responsible for maintaining the analytics.

When embedding, modern analytic solutions offer more than simple iFrames with their limitations and security risks. They allow the embedding of analytics using APIs and scripting to embed anything, anywhere. Look for platforms that are cloud-delivered SaaS applications with robust, flexible embedding options.

What was modern five years ago isn’t sufficient for today’s analytical applications. Needs have changed, data has multiplied, and technologies have advanced. Keep these things in mind when choosing a platform for your analytics. You don’t want to get trapped implementing a system that, while appropriate in its day, is no longer ideal for the new world. Business-friendly, flexible, self-implementable, and efficient in the movement of data are all features that are readily available today. It’s worth it to sort through the many old-style options to find those that will ready your business for the future.