We get this question a lot, and why wouldn’t we? SaaS companies are usually shopping for BI on a deadline, looking for a solution that will get them to market with the least amount of effort and in the shortest time possible.

But the length of time it takes to implement an embedded business intelligence application varies widely from case to case and hinges on dozens of variables, like security requirements and the size, type, and number of data sources, for example. So when we get this question, we usually say “it depends,” and it wasn’t until recently that we decided we could do a better job of articulating what that means.

So last month, we reached out to our clients with a survey asking for detailed information about their implementation of Exago BI. What follows is a summary of their experiences and the feedback we received.

So How Long Does it Take, Overall?

Though nearly half of the companies completed their implementation in under 2 months, it took our clients anywhere from 2 days to 12 months to build a functioning implementation of Exago BI.

The company reporting the shortest implementation time (2 days) based their answers on the time it took to set up a test bed during their evaluation of the product. This was a speedier process for them because they only ever intended to use the application as a reporting engine, not for ad hoc report design. They didn’t, for example, customize the UI in any way or need to set up report security.

At the other end of the spectrum, one of the companies that required a year to fully implement the application spent the majority of that time (6-9 months, due to the volume of data) building a .NET assembly data source, a business object layer that would sit between Exago and their database. An assembly layer facilitates advanced data source customization and, in their case, serves as a central means of managing data access, row-level security, business logic, and user authentication. Because of its complexity, the layer required more time to architect but has afforded the company greater flexibility in the long run.

In the middle of the pack, by contrast, a company that implemented Exago in 3 months reported needing only seconds to connect to their data sources but required a month to define those sources.

What Took the Longest?

The majority of respondents reported that preparing their data sources for reporting was the most time consuming part of the implementation. This preparation might include creating a .NET assembly, as in the example above, or formulating joins between objects to create a data framework. Since companies need to prepare their data for reporting regardless of which BI platform they use, this statistic suggests that the majority of Exago users spend the most time on processes exclusive of the application. Taking the time to optimize your data is generally considered good practice.

Report building came in as the second lengthiest implementation process. According to the survey, building a report can take anywhere from minutes to weeks, depending on the report’s complexity. In some cases, a very advanced report might require the development of a programmatic extension, such as a custom function or a server event. In other cases, it might require the creation of a lengthy template or a new set of data objects. It can be near impossible to anticipate these types of case-specific needs until they surface during the report-creation process, lengthening the time it takes to develop a report repository for end users.

The data analysts creating this repository must also be trained in the software, which adds to report creation time as well. As one respondent writes, “The tool is great, but not everyone knows how databases work…[or] how to [get] the right data on the report.” This learning curve also factors into implementation time.

Tied at the bottom of the list are API integration and security, both of which can become more time consuming if the situation calls for unorthodox arrangements. One company, for example, chose to implement the API in a way that would store report definitions in a database, and while they report API integration as one of their most time-consuming processes, they also concede that this was, in part, because “the API and customization options are extensive and take time to understand and work into an existing framework.” They go on to add that “the documentation was sufficient, and support was responsive to questions and issues during the implementation process.”

What Resources Did Companies Utilize?

Almost all respondents made liberal use of the Exago Support Center, and most used provided code samples as a springboard. Half made use of Client Services, a comparatively recent addition to Exago’s deployment offerings, and the remaining companies reported either using the new Demo Application or corresponding with particular individuals back when many of these other resources had not yet been created.

Conclusions

Overall, feedback on the implementation experience is positive. One respondent writes that “Exago was very helpful and responsive. Their product provides a lot of opportunities for extensibility which allows users to customize the implementation quite a bit.” Another speaks to the ease of deploying Exago for evaluation purposes, explaining that “other products took too much effort or resources” to install for testing by comparison.

From the results of this survey, we can generally conclude that it takes about 4 months on average to implement Exago with most of that time going toward data preparation and report design. The Support Center has proven to be the most valuable deployment resource, followed closely by code samples, suggesting that we should continue to build out these resources and work to make others more accessible.

So while the time it takes to implement Exago does still hinge on each company’s context and requirements, we hope having a ballpark timeframe will help with your release planning.