Every business needs to make some sense of the key metrics and data that drive their business. Traditionally, a business might get started by buying a reporting tool and building some reports using the production database of their core business applications. This might work for a while, but soon this approach runs out of gas due to:
1) Difficulty of Report Writing – It’s actually not that easy to write reports against a database designed for transaction processing. The database is just not designed for ease of reporting. Sure you can do it, but you’ll find that you’re in endless loops where the business users try to describe what they need, and the developer keeps falling just short of what the business wants.
2) Performance – Eventually you’ll find that you are harming your application’s performance every time someone runs a complex report.
3) Lack of an enterprise wide view of data and metrics – It’s unlikely that all of your company’s key data is located in a single database. So if you want an enterprise-wide view of your business, you can’t create reports that tell this story.
So what you need, then, is a Business Intelligence (BI) system that gathers up data from disparate systems across your business, and creates a data warehouse that’s built for ease and efficiency of reporting (not for transaction processing efficiency). And you need some nice tools for creating reports on top of this data, and for distributing it to the web and mobile devices. Crucially, this lets your smartest analysts build their own reports, rather than depending on software developers to do it for them. This is both more cost effective, and more satisfying to the analysts, who would rather do this work themselves rather than depend on a software developer for every tweak to a report.
BI tools aren’t new, but traditionally they’ve been very expensive. Depending on the size of your company, you could easily spend six figures on the BI tools themselves. And then you’d have to spend a similar amount, or likely a multiple of that, for BI consultants or internal resources to build the data warehouse and the initial reports and dashboards.
Luckily, open source BI tools have emerged in the past few years, and you should consider them as a strong alternative to BI tools such as those provided by Oracle, IBM, or SAP. As an example, Pentaho is a vendor that offers a community edition of their BI suite (in addition to an Enterprise edition). You can use the community edition for free. Pentaho offers a complete BI suite including ETL tools to extract your data and build the data warehouse, report writing tools, dashboard tools, and a web portal for distributing the reports to users. And you can use an open source database like MySQL instead of paying a vendor such as Oracle for that.
OK so we’ve cut out hundreds of thousands in licensing costs. Now what about implementation costs? At Cayugasoft we use primarily offshore resources for this, located in Eastern Europe. The costs are significantly less than equivalent US-based resources, and in our experience, the results are the same. And as opposed to a BI engagement with a US-based consulting firm, you might find you want to keep some resources on your project (at a lighter level) to continue to write reports for users. Using offshore resources, you can afford this — you don’t need to chase the consultants out as soon as you can.
On Cayugasoft’s website, you can find a case study that shows an example of a dashboard and a drill down analysis tool built using Pentaho, and some open source data on US imports and exports. Please look at the case study and reach out and contact us to discuss what our approach to open source BI tools can do for you.