Choosing an embedded analytics solution can be an intimidating task. Get it right, and you’ll add huge value to your application, cementing its place with your existing customers and increasing its saleability for many years to come. Get it wrong, and you’ll waste a huge amount of time, effort and no doubt money. So, we thought we’d put together an article to help guide that decision. It’s biased, obviously, we are after all an embedded analytics vendor. But it will, we hope, ring true for any who have recently gone through the process.
First off, any solution you choose should look great. Analytics is first and foremost about visualising data so it can be understood at a glance. If a vendor can’t get that right, walk away. You should also be able to completely white label it. Embedded analytics is about improving your product, not selling someone else’s.
Second, any embedded analytics solution should be easy for the developer to use and integrate. Seamless integration with the business application needs data to be prepared at the back end, and the visualisation components to be embedded and integrated at the front end. If either of these tasks require the developer to learn a new language, or significant new syntax, it defeats the object of licensing a third-party product.
DataPA OpenAnalytics allows business logic to be expressed in a variety of languages, such as SQL, .NET or OpenEdge ABL. Moreover, our technology allows developers to hook into existing business logic already developed for the business application in any of these languages, to provide data transformation in the analytics layer. After all, why write it twice? We also make sure our training courses and documentation are tailored to specific audiences, so the examples and syntax can be lifted directly into our customers code. We don’t expect our customers to embark on weeks of training before they are comfortable developing with DataPA OpenAnalytics, and we don’t think you should consider a vendor that does.
The power of embedded analytics is providing intelligence within context. An example might be providing a sales person with information about past purchases for a product at a glance, as and when they are making a sale. To deliver this, it’s key that the analytics layer and business application are tightly coupled, with information flowing in both directions between the analytics layer and business application. This means it’s important the analytics engine has a flexible and complete API and component set, so analytics content can be tightly integrated in any client application, be it web, mobile, client, even wearable devices. Above all, it should be easy for the developer to leverage the analytics capabilities on every platform they develop for.
Delivering analytics is about more than just data visualisation. There’s a huge array of extremely impressive open source data visualisation components available today, free and simple use. If you just want to visualise simple data, why pay for a full analytics solution? To add value over and above these free components, an analytics solution must make it easy to draw data from anywhere and then blend, accumulate and transform that data to present it in the visualisation layer. It’s this analytics engine that makes the difference between simple data visualisation that allows for a pretty picture in a demo, and something that will deliver real, practical intelligence to an organisation.
Finally, and perhaps most importantly, make sure the company you choose to provide your embedded analytics combines real expertise in analytics with agility and a drive to innovate. As we’ve discussed previously in this blog, the analytics sector like most software markets is on the verge of huge change driven by the developments in big data and machine learning. Choose the right vendor and this disruption offers huge opportunities for your business and your customers.