Does Hand Coding Unnecessarily Add to the Cost of Data Integration?


Last week, I wrote a post on Why Companies Still Hand Code, But Don't Have To, despite the numerous data-integration tools of all price ranges available on the market. In the piece, I summed up the reasons cited by Rick Sherman, founder of Athena IT Solutions, which included ignorance of market costs, a lack of resources and a misplaced adherence to corporate standards.


Several readers commented on the post, adding that Sherman overlooked one critical reason why so many data-integration projects are hand coded: The human factor.


When it comes to data integration, the missing human factor is simply that developers like to code. They're comfortable with the tools and, in that moment, it's easier to hand code than evaluate, choose, buy, install, learn and use a new tool-even if that tool might shave months off projects on down the line.


It's a bit ironic, really, because IT folks are always complaining about user error and the human factor mucking up their projects, and yet, when it comes to data integration, they're making a pretty human mistake: Focusing on their own short-term needs and comforts rather than seeing the bigger picture.


To tell you the truth, I'm certainly not going to say they're wrong. For all I know, these tools could end up costing more in the long run and be just as much trouble as hand coding.


But that's not what the experts seem to think. They seem to think tools could save money and time on data integration.


If you're skeptical about whether data integration tools can save you money, you might want to check out the ROI calculations in this recent Talend white paper, "The Return on Investment of Open Source Data Integration." It looks at how hand coding, proprietary data-integration tools and open source data-integration tools fare in three types of data-integration projects-a small, easy project; a medium project and a large project with costs running into the millions.


Now, you can probably guess which approach is going to come out ahead just by reading the title and knowing that Talend is an open source data-integration vendor. However, it's still interesting to see how it breaks down the costs and just how expensive hand coding becomes when you're doing a large project-even when it's compared to proprietary solutions. In one scenario, the costs for hand coding is double the cost of a proprietary solution and four times the expense of an open source solution over a five-year period.


If nothing else, it will help you to think more broadly about how you're determining ROI on data-integration projects. When it comes to choosing a data-integration tool, IT's very particular, doing the research on different solutions, taking a close look at the long-term cost of ownership, examining how it supports governance, and so on.


But are you holding hand coding up to the same rigorous evaluation? Probably not. And that could be costing you more than you know.