Grants.gov

As many know, I have been developing grants management systems for the federal government for years. I'm particuarly disappointed to read a GAO Report released yesterday on Grants.gov. The problems facing this system are significant. I am particuarly concerned about the funding aspect. Do you mean to tell me that with about $18 Million per year we can't make this thing work? Even if some agencies pay late, it still looks like about $12 Million more than what you should need to do the job.

Thinking back I remember some of the problems that we had in the form and format of the output we received from Grants.gov. It came as basically ascii without any format at all. But to be what, 4 or 5 years in and still be working through fundamental problems like DUNS registration, that is not acceptable.

What is the fundamental problem here? The problem is that we have a bunch of agencies who have no inclination to work together and have evolved different internal processes for managing their grant processes. I know that at least two partners have radically different methods for handling it. But what does Grants.gov try to do? They try to squeeze every peg through the same round hole. That is a fail whale right from the get-go. Instead of trying to get each agency to accept a consensus process, carve it up a little bit. Let HUD have their little unique aspect, and let USDA have theirs. The opportunity that Grants.gov should focus on is having access t each of these on the same system, but each agency will have a slightly different application. Grants.gov should focus on the development of a standard base model that can easily and efficiently be customized based on nuance with the agency or the program.

The cost model seems to be a little bit off doesn't it? Instead of a flat rate that Departments and Agencies pay in, wouldn't a utelization-based cost model be better. What happens in this current model is that agencies pay their burden and then over-utelize resources. That second tier is where they are putting a lot of effort and that effort, I bet, isn't in proportion to the amount paid by the agency. But this would be a difficult case to make. So instead, the Grants.gov PMO should focus on the base model, which is deployable for basically nothing, and charge based on the amount of customization on top of the base and the tier of support the agency wants. So if you have a program that will collect applications and you need a lot of support for the community then you pay more. If you have a simple program and all you are collecting is basically an SF 424 and a lobying certification, and you don't need help, then you pay less.