Embedded Analytics

The Risks of Cutting Corners: Choosing the Right Analytics Evaluation Approach

By Alexandra Thrasher
Share on LinkedIn Tweet about this on Twitter Share on Facebook

The concept of “testing software” has different meanings for different application teams. And the method you choose can have a big impact on whether your embedded analytics solution succeeds or fails.

One rushed decision can impact your organization for years. “Once you have embedded analytics into your application, it is a major effort to switch to another provider,” writes Gartner in their report, 5 Best Practices for Choosing an Embedded Analytics Platform Provider. If you choose the wrong analytics solution and need to rip it out in a year or two, it can incur technical debt that may never go away. What’s more, you’ll have to go back to the beginning and face a much longer cycle to successfully embed analytics.

>> Related: How to Break Up With Your BI Solution <<

The best way to avoid a costly rip and replace is to choose the right analytics solution from the start—and that means taking time to evaluate the top vendors on your list. Here are the three most common evaluation approaches for embedded analytics:


Free software trials are common—but they’re a double-edged sword. In a trial, you have total control to do what you want with the platform. You’re completely let loose. If you’re like 90 percent of other software teams, you’ll probably create some basic analytics and maybe a dashboard or two. That leaves you feeling good about the solution.

However, what goes unevaluated are the more advanced requirements or the ones that are on the horizon in the next couple years. Free trials stop short of providing an accurate picture of future success.

Proof of Concept (POC)

In a typical proof of concept approach, the analytics vendor creates a POC project for you. Most vendors will recreate what you have now, maybe include some additional features, and demonstrate the result.

But unless it’s part of a larger evaluation, a POC will leave you with unanswered questions: How easy was it to create the project? Will your development team be able to do what they want in the platform? What specialized expertise was used? Who knows—but the POC looks great.

Structured Evaluation (also known as a Technical Evaluation)

A structured evaluation is more thorough than a trial or standard POC, because it includes criteria you have now and criteria you may not even be considering. This approach empowers product teams to:

  • Test technical requirements and sophisticated capabilities in their environment
  • Evaluate a vendor’s approach to customer success
  • Familiarize developers with the platform, so they can get up and running quickly
  • Get to market quickly, since it does not require a lot of time or resources.

Trials and POC approaches have their place in some software evaluations. But for an embedded analytics project with sophisticated requirements (literally anything beyond a basic data visualization or interactive dashboard), the easiest and most effective way to assess a vendor is a structured evaluation.

It’s true that a structured evaluation takes a little more time and effort than a standard trial or POC. But that’s nothing compared to the pain of choosing the wrong solution and having to rip and replace it in a few months.

To learn more about choosing an analytics vendor, read our BI Buyer’s Guide >


Originally published March 26, 2019

About the Author

Alexandra Thrasher is a customer content program manager at Logi Analytics, where she partners with application teams to get the most out of the Logi platform. Prior to Logi she was a solutions consultant at LinkedIn.