Depends. I was familiar with http://www.informatik.uni-trier.de/~ley/db/indices/a-tree/a/... and to be successful, it has to be a great product. I.e. find where the pain is and solve it. Success lies on the product side, not the research...