‘Owning’ a business intelligence (BI) solution is no longer the only possibility for a company striving for informed decision-making. Data analytics outsourcing market is forecasted to grow, which means that more and more companies will choose ‘renting’ their BI over any type of implementing it.
As a provider of both BI as a service (also known as managed analytics services and data analytics outsourcing) and BI implementation services, ScienceSoft articulates the main difference between these alternatives.
- BI implementation presupposes a 6-8 month project; once it’s completed, a company gets a data analytics solution (whether on-premises, in-cloud or hybrid one).
- BI as a service presupposes continuous cooperation, where the outsourcing partner is totally responsible for the solution’s technical embodiment, and the company gets access to the web interface with insightful reports already after 6-8 weeks.
Let’s look closer at the approaches to BI as a service from the perspective of analyzed data, which we have singled out based on our experience.
BI as a service based on internal data
Relying solely on the data retrieved from a company’s systems, this approach provides insights into the company’s business processes, customers, performance, and more. The main advantage of this approach is that a plethora of data is usually available for the analysis. At the same time, the company is limited to their ‘internal wisdom’ – they lack external data to compare themselves against.
Here’s a project from our practice to illustrate the approach. One of our customers requested us to help them prioritize their product categories as they wanted this strategic decision to be data-driven. The project was based on their 2-year financial and production history. During the project, we scrutinized the data to find meaningful insights, and built reports and graphs so that the customer could glance at the findings and answer the business question they had.
BI as a service based on external data
This approach is frequently used to run strategic market analysis or social media analytics. The biggest challenge while working with external data is ensuring data quality, as it usually comes from multiple disintegrated data sources and can contain errors and discrepancies.
Here’s another project from our portfolio. A management consultancy outsourced the obtained market data to us to get valuable insights into different industries in various economic conditions. To solve this task, we deployed BI infrastructure with a data warehouse and a 40+ dimensions OLAP cube on our server, while the customer was provided with the access to the pre-built reports and dashboards, with the possibility of running ad hoc analysis.
Using data from both inside and outside, a company can get the broadest picture that encompasses both their internal operations and the market perspective. No wonder that a hybrid approach is the most widespread.
According to Business Application Research Center, companies mostly derive data from multiple data sources. Besides, there’s a trend to increase the number of data sources (a half of BARC’s respondents believe that they already experience this trend). Also, the number of data sources naturally depends on the company size: BARC names a median number of 5 internal sources for mid-sized companies and 10 sources for large ones.
Is there the best approach?
Of all the approaches, we consider the hybrid one the most efficient as it ensures the variety of data sources and offers better scope for analysis. However, the scenarios with getting insights from just internal or external data are also possible, as we can see from the real-life examples described.