The appointment of James A. Hofbauer to the position of President and CEO of CLSI and the departure of Robert Walton (on October 9, 1992) have triggered many questions about the company: Is it in trouble? Is the company's product functionally competitive? What are the risks if the company ceases to develop and support the product?
We've hesitated to write about this topic lest it add to the company's problems, but there now are so many rumors that a straightforward presentation of the facts is warranted. In this article we seek to provide such information as is available at this time, most of it drawn from published articles, press releases from CLSI and other companies in the industry, and consultations with CLSI clients.
The company's financial condition appears to be of the greatest concern to librarians. (Perhaps they remember the predecessor company, Computer Library Systems, Inc., AKA CLSI, founded in 1971, went through bankruptcy in 1979 and a successor company, CL Systems, Inc., which was also known as CLSI was set up to continue to sell and service the CLSI product line.) By 1985, with the name officially changed to CLSI, Inc., the company was the leading library system vendor and was purchased by TBG, formerly Thyssen-Bornemisza, a privately held, multi-national corporation with annual revenues in excess of $1.7 billion. This association was hailed at the time as a way of providing additional resources with which to grow the business. However, even with the difficult financial times in the 1970s, the past three years have in some ways been the most difficult because there are now other strong competitors ready to take advantage of any weakness.
The general perception is that CLSI's sales have been poor the last three years. An examination of CLSI's responses to the annual surveys conducted by Library Systems Newsletter are revealing in this regard: see Table 1.
Gross revenues are more difficult to track. CLSI reported sales in excess of $20 million for 1984 and 1985; in excess of $30 million for 1986 and 1987; $35 million in 1988; back to $20 million in 1989; up to $30 million in 1990; and, back down to $20 million in 1991. The company has not claimed to be profitable in the past three years; however, given the substantial income from maintenance and upgrade sales, any losses would have had to be relatively small. Considering this, along with the financial strength of the parent company, TBG, it would appear that the financial condition of CLSI should not be the foremost concern.
Another perception is that staff turnover has been high. Anecdotal evidence suggests that more than two-thirds of the professional staff have left in the past three years. Even more critical, the company has had four presidents in that time: Gene Robinson, James Barrentine, Robert Walton, and James Hofbauer.
Declining market share and rapid staff turnover have been obvious but software development also has been problematic. Table 2 shows the major modules and submodules that were committed to four academic and public library clients between 1986 and 1990, along with the promised dates of delivery. At the beginning of 1992 the libraries were still waiting.
The new online patron access catalog module-now renamed LIBS 100plus-was finally made available in general release on June 24, 1992. Although CLSI has recently advised us that disk update logging, report generator, support for x.25 packet switching, and electronic mail are part of the LIBS 100plus release, as of mid-October 1992, none of these libraries appears to have received any of the functionality committed except for MARC tape output. Authority control with global change is now scheduled for 1993, as are acquisitions and serials control.
Several of the libraries in Table 2 have been advised that they would not be able to load the new online patron access catalog module-now renamed LIBS 100plus - unless they upgraded the system at a cost of up to $100,000. Several had contract clauses which stipulated that the vendor was obligated to deliver all hardware necessary to deliver the functionality. After considerable discussion some of the upgrade costs were reduced to as little as $20,000. However, even after the hardware upgrades are completed, the libraries will still have to wait for most of the functionality, including the new acquisitions and serials control modules committed for delivery as early as 1990 and 1987, respectively. The total number of CLSI accounts which had signed contracts for upgrades to LIBS 100plus by late September 1992 was 26.
Now that CLSI has migrated away from its proprietary operating system to UNIX, there is less risk for libraries which install the product and subsequently cease to have support. The Sequent and IBM RS/6000 hardware platforms used by CLSI are supported by several other automated library system vendors, as is the operating system. Most major vendors have successful experience migrating files from CLSI systems to their systems. (Dynix has migrated at least 17 accounts; DRA, Gaylord and INLEX at least two each.) Nevertheless, it would be a costly proposition because there may be some hardware reconfiguration required, the operating system probably would have to be relicensed, new applications software would also have to be licensed. Furthermore, training and other service costs would have to be paid. The cost of changing to another vendor could be as much as 50 percent of the initial outlay-even for libraries that purchased a new system only a few months ago.
Table 1. A Compilation of CLSI Responses to the LSN Annual Surveys of Library Automation Vendors, 1983-1991. (* Not Supplied)
| Year | Total Number Installed Systems | New Name System Sales | Number of Staff Software Development | Number of Staff Customer Support |
| 1991 | 309 | 1 | 28 | 80 |
| 1990 | 335 | 9 | 79 | 85 |
| 1989 | 345 | 19 | 34 | 67 |
| 1988 | 331 | 42 | 89 | NS* |
| 1987 | 302 | 35 | 79 | NS* |
| 1986 | 277 | 36 | 100+ | NS* |
| 1985 | 259 | 32 | 81 | NS* |
| 1984 | 235 | 26 | 61 | NS* |
| 1983 | 209 | 29 | NS* | NS* |
- There has been a substantial attrition rate in the customer base. From 1983 through 1991 CLSI reported selling 229 systems, but the net gain in accounts was only 100; therefore, 119 systems were unplugged during this time. The attrition rate was particularly bad from 1987 through 1991-a period during which 71 new systems were reportedly sold, but the total number of installations declined by 22..
- New system sales reached a peak in 1988, with 42 systems sold. From 1989 through 1991 only 29 systems were sold, including a low of only one new system in 1991. The company has continued to make major upgrade sales to its existing customer base at the rate of two per month..
- Software development staff figures have fluctuated a great deal, with the 28 staff reported in 1991-the lowest in nearly a decade..
- Customer support staff (the post-sales staff) has remained relatively stable-down only five people in the past year.
| Library 1 (1986) | Library 2 (1987) | Library 3 (1988) | Library 4 (1989) | |
|---|---|---|---|---|
| Online Cat (PAC II) | At delivery | 4/1/1988 | 10/1/1991 | |
| Authority Control | At delivery | 10/1/1989 | 10/1/1991 | |
| MARC Tape Output | 1/1/1987 | 10/1/1988 | 7/1/1992 | 10/1/1989 |
| Disk update logging | 1/1/1987 | |||
| Serials Control (new) | 1/1/1987 | 7/1/1990 | 10/1/1991 | |
| OSI Support | 1/1/1990 | Within 24 mos. of publication | 10/1/1990 - for record transfer | |
| Acquisitions (new) | 4/1/1990 | 10/1/1990 | ||
| Interlibrary Loan | 10/1/1990 | 10/1/1990 | 18 mos. from Stds. Pub. | |
| X.400 Support | 7/1/1989 | 12/1/1990 | 10/1/1990 | |
| Report Gen. (new) | At delivery | 10/1/1989 | 10/1/1990 | |
| Journal Citation | 7/1/1990 | 7/1/1991 | ||
| Community Information | 4/1/1991 | 10/1/1991 |
Under what circumstances should a library consider purchasing a CLSI system? The most clear-cut situation we can envision is a competitively bid procurement in which the CLSI price is so low that even the subsequent cost of migrating to another vendor's system would not make the total outlay significantly more than the cost of purchasing another system in the first place, or if the promised and undelivered functionality is of little or no importance to that library. Even at a super-low price, a library should recognize that it is purchasing a product which is functionally less complete at this time than a number of other products on the market. DRA, Dynix, Gaylord, Innovative, and NOTIS-companies with a substantial share of the market-- all have the core modules of acquisitions, serials control, circulation, and online patron access catalog in general release, as well as a number of additional modules.
Saying this, we remain hopeful that this venerable company will once again rise to the challenge.
[Contact: CLSI, Inc., 320 Nevada Street, Newtonville, MA 02165; (617) 965-6310.]
