Open Source OPAC Market Penetration-beta 2

Bob Molyneux writes: This post presents a revised version of the summary numbers that appeared here on October 15 analyzing use of open source OPACs in U.S. public libraries. This is the second beta, if you will, on the fuller report that I hope to start doing regularly.

A discussion ensued at Web4Lib and Joshua Ferraro, President of LibLime asked for a more detailed explanation of how I got my numbers since some differed from lib-web-cats, my main source for identifying which libraries used which software. I answered at some length and those interested in a detailed discussion can go there. The earlier post dealt with the same issues, of course. I will summarize the longer Web4Lib note here.

Bob Molyneux writes: This post presents a revised version of the summary numbers that appeared here on October 15 analyzing use of open source OPACs in U.S. public libraries. This is the second beta, if you will, on the fuller report that I hope to start doing regularly.

A discussion ensued at Web4Lib and Joshua Ferraro, President of LibLime asked for a more detailed explanation of how I got my numbers since some differed from lib-web-cats, my main source for identifying which libraries used which software. I answered at some length and those interested in a detailed discussion can go there. The earlier post dealt with the same issues, of course. I will summarize the longer Web4Lib note here.

I used what I called the Missouri method;that is, show me to classify the libraries. So, when I say the library is running Koha, I have checked and seen it on its Web interface. lib-web-cats uses a less restrictive method. I am uncertain exactly how to describe it accurately but I believe if the library announces it is switching vendors to (say) Koha, lib-web-cats lists Koha as the vendor. Koha actually runs on 15 systems that I have seen. lib-web-cats credits an additional 26 of which 25 are still running their legacy software, that is, not Koha. One does not have a Web-presence and I concede its exclusion is dubious.

We have, then, two different methods;each reasonable and explicit for classifying these libraries and the analyst will do both and bracket the true but unknown numbers. However, 23 of these systems credited to Koha and running legacy software are members of INCOLSA. INCOLSA announced a partnership with LibLime and you will note down the page a bit that LibLime offers support for these libraries for Koha and Evergreen. So, which is it? When the dust settles, which libraries will run which one?

That was where I was stopped with the first beta on October 15. While replying to his post, I had the idea: why not add a line with these libraries included and give two totals? So, here 'tis:. Table 1 is the same as the table from October 15 and Table 2 is the new one with the less restrictive classification of open source public libraries. I report; you decide.

1. What percent of U.S. public libraries currently run open source OPAC software? Some figures:

Table 1: Summary data on U.S. Public Library OPACs (Missouri method) (same as 10/15/2007)

Table 1

Vendor Systems Population
served
Total
circulations
Total
expenditures
All U.S. Public Libraries 9,207 286,730,444 2,010,777,017 $8,643,027,806
Koha 15 235,755 2,383,624 $7,203,945
Evergreen 47 4,470,670 19,073,650 $66,242,856
Totals for open source (Missouri): 62 4,706,425 21,457,274 $73,446,801
Open source
as a percent of total
by variable:
0.67 1.64 1.07 0.85

Table 2: Summary data on U.S. Public Library OPACs (less restrictive method) (new)

Table 2

Vendor Systems Population
served
Total
circulations
Total
expenditures
All U.S. Public Libraries 9,207 286,730,444 2,010,777,017 $8,643,027,806
Koha 15 235,755 2,383,624 $7,203,945
Evergreen 47 4,470,670 19,073,650 $66,242,856
Other 26 440,596 6,044,984 $18,197,334
Totals for open source (unrestricted): 88 5,147,021 27,502,258 $91,644,135
Open source
as a percent of total
by variable:
0.96 1.80 1.37 1.06

As can be seen, the totals do not change much because the new libraries are all small with the exception of Howard County, which totals I gave on October 15.

This time, I produced the data in a spreadsheet if anyone wants it. I included five tables with these data arranged in various groupings. One used the unrestrictive method and had all Koha libraries by that method. There are 41 systems. Howard County continues to stand out. It has 39% of the population served of these 41 systems, 56% of the circulation, and 46% of the total operating expenditures.

The original datasets that these spreadsheets are derived from are in SAS format. I then used the Open Office Calc for a spreadsheet and will use it to convert the files to .xls for anyone who wants it that way. I am uncertain of how well this will translate—this one would be a third generation copy—but it usually seems to work OK. If you want the SAS files, let me know.