What’s the value proposition of a pre-built analytic application?

In recent posts, we’ve examined various ways that companies seek to get faster returns on their investment in BI, from better defining the value proposition of BI to using software as a service (SaaS/cloud) BI applications. Another method we see a number of companies explore is leveraging pre-built analytics. Pre-built analytic applications are designed to be “business intelligence in a box”. They contain extract ETL, data warehouse models, cubes or other reporting structures, and reports and interfaces.

In short, the main value proposition for pre-built analytic applications is that have the potential to be cheaper, faster to implement, easier to maintain, and more feature rich than custom built analytic applications. However, in the wrong environment they can be impossible to maintain, more expensive, and unable to adapt to an organization’s unique needs. As business intelligence consultants our clients often ask whether or not a pre-built analytics solution makes sense for them. As in all make vs. buy decisions, the existing systems and requirements play a key role in the decision. Here is a framework we have developed and used with clients to help them make the selection:

BI Build v. Buy Criterion #1: Number of Sources

RULE: More than one source system increases the need for custom integration.

Choose Custom if:	Choose Pre-Built if:
Multiple sources. Production, sales, and or marketing in different systems. Multiple business lines require integration: either across products, or across geography. As a result of major mergers, different units have different business processes.	Single major source for all key data. Integrated product lines/single geography.

BI Build v. Buy Criterion #2: Source System Customization

RULE: More than one source system increases the need for custom integration.

Choose Custom if:	Choose Pre-Built if:
Source is a significant customization of packaged software or fully custom system. System not always at the package supported version. Multiple instances of the same major system with different configuration, version, or customization.	Single plain-vanilla system running current vendor version.

BI Build v. Buy Criterion #3: Source System Complexity

RULE: Extracting from big ERP (SAP, Oracle) is hard, pre-built extracts add value.

Choose Custom if:	Choose Pre-Built if:
Not Big ERP Highly customized ERP. Have already built data extract	Extremely complex source systems. Big ERP. Do not have table and data level documentation.

BI Build v. Buy Criterion # 4: Integration Requirements

RULE: Complex integration will require custom work.

Choose Custom if:	Choose Pre-Built if:
Several “best of breed” sources.	Key data (sales, customer, product, marketing, orders, inventory) in a single system. Have existing integration processes for real time and external coordination.

BI Build v. Buy Criterion #5: Vendor relationship

RULE: pre-built applications require high levels of vendor commitment

Choose Custom if:	Choose Pre-Built if:
Concerned about single-vendor account control. Limited negotiating leverage.	Committed to a single vendor. Strong vendor partnership.

BI Build v. Buy Criterion #6: Cost and Resources

RULE: Costs depend on existing environment and previous choices

Choose Custom if:	Choose Pre-Built if:
Have existing BI/DW team and infrastructure. BI/DW team has predictable deadline delivery and cost structure.	Looking to jump-start a BI/DW effort. Aggressively minimizing IT head count.

BI Build v. Buy Criterion# 7: Importance and Competitive
Advantage

RULE: Key processes benefit from customization, standard processes benefit from pre-built.

Choose Custom if:	Choose Pre-Built if:
processes are unique and company specific: (e.g. Sales, Marketing, Production)	processes are regulated or standardized: Finance, HR, Inventory.

Bottom Line on Pre-Built Analytics Applications:

So which option works best for you? In a nutshell, the decision boils down to the level of simplicity and standardization of your existing information and infrastructure. Of course, as we discuss in a number of other posts about developing your BI strategy, this build vs. buy framework is only part of a larger framework for designing, implementing and leveraging a business intelligence that increases company profitability.

by Tom Victory, Principal Consultant

One Response

1. 5 Common Mistakes in Implementing Pre-Built Analytic Applications | Decision Path says:

   May 31, 2011 at 9:41 am

   [...] our previous post on pre-built analytic applications, we described a framework for choosing between custom and pre-built business intelligence.  This post describes common pitfalls in implementing pre-built [...]

   Reply