[ This post is authored by David Dorman, US Marketing Manager for Index Data. David responded to an email I sent to marketing contacts at commercial and open source providers of federated search products and services asking them to engage in my campaign to create publicly accessible demo applications with identical connectors to facilitate customer evaluation of offerings. With David’s permission I am publishing his response in its entirety. I will post my response to his letter in the near future. David’s reference in the first paragraph to the “Metasearch Smackdown” is to this Library Journal blog post by Roy Tennant. David can be reached through the contact information for Index Data provided in this blog’s vendor information page. ]
I’ve looked over the points of the metasearch comparison campaign and it seems like something Index Data would like to explore participating in. Have the databases been selected yet? I do, however, have some reservations about two assumptions that you and Roy seem to be making with the Metasearch Smackdown idea.
1. One is that each vendor has a “product” that can be compared to all the other vendor’s “products.”
This is not a valid assumption for Index Data’s metasearch offerings. It is true that our basic MasterKey hosted service is a product that can be compared to the other vendors offerings. However, Index Data offers to customize or extend the functionality of the MasterKey interface to meet any clients requirements if MasterKey does not already meet them. This business practice will not be apparent in any use test.
Index Data also distributes its core metasearch engine, Pazpar2, under a GPL, so that any library can freely make use of it. It would take more than the proposed evaluation methods to evaluate the capabilities of Pazpar2, and the opportunities it offers libraries.
2. The second assumption is that each vendor’s product is its product alone.
In the case of Index Data, MasterKey can use both Z39.50 connectors and WebFeat (now Serials Solutions) translators. Your smackdown does not take into account that two different methods of connecting to the same databases can be selected by a MasterKey customer, if those databases have Z39.50/SRU compliant servers. If the customer chooses to use Serials Solutions translators, then the product is a combined Index Data/Serials Solutions product. If the customer chooses to use only Z9.50 connectors, then the product is totally an Index Data service. There are a lot of potential permutations of MasterKey and Serials Solutions that could be set up with those 10 sample databases, and separately evaluating which way of connecting to a particular database works better would be pretty involved process.
If only comparing metasearch services were simple! But the reality of the market is just too complex for easy and clear comparisons, and for better or worse, that complexity, and the concomitant flexibility that is its rationale, will just keep on increasing.
One final note: I neglected to mention in my previous email to you that ever since the release of the MasterKey hosted metasearch service in June 2007, Index Data has made the use of it freely available to anyone at http://masterkey.indexdata.com. So anyone can easily see for themselves how the basic MasterKey hosted service works.