I. INTRODUCTION
In its recommendation to the LMS Steering Committee on April 30th of last year, the LMS Evaluation Committee warned:
“We must not take the current inadequacies of the [Voyager] system’s technical services components lightly. Even with the improvements in the ’98 releases, we should expect severe delays in production that are likely to continue for some time, and CUL will need to factor these delays into the cost of implementation. While it will be incumbent on technical services departments to re-examine their respective workflows to adjust to the new LMS, the Library must also be prepared to: (1) tolerate processing backlogs that are likely to develop, [and] (2) provide additional staff to technical services units -- as many as 6 FTE for up to 3 years ... These estimates can be mitigated if Endeavor continues the integration of its technical services modules and continues to make their use more efficient.”
What follows represents a first crack at a systematic prediction of the effects of Voyager ’99 and 2000 implementation on CTS Acquisitions and Cataloging workflow and suggestions for adjusting that workflow to meet our needs in the new environment.
This is an informal discussion paper. By that I mean that we should consider it as a series of assumptions and arguments to be probed and questioned, corrected and expanded, and otherwise fine-tuned. While some of the information on which the following assumptions and projections are based is well-established, other facts are less precise and more intuitively derived. Obviously, the details of workflow in the Voyager environment and the pace of that workflow are especially difficult to predict and estimates in this latter regard will be particularly impressionistic.
The following assumptions pertain to the entire study:
1. New acquisitions have remained at FY 1998/99 levels and will continue to do so for the post-implementation recovery period.
2. For the first 52-55 weeks on Voyager (i.e. through FY 2000/01), it will take longer to carry out basic functions. The following estimates will apply:
Month 1 ... a 25% decrease in current productivity
Month 2 ... a 10% decrease in current productivity
Month 3 ... a 15% decrease in current productivity (includes upgrade to Voyager 2000)
Months 4-12 ... varies according to function (see individual entries in Sections II & III)
Months 1-4 (composite) ... varies according to function
FY 2000/01 (composite) ... varies according to function
I’ve referred to these percentages as the “Voyager factor” in the analysis below.
3. Assuming that we implement and assist in the development of Gary Strawn’s tools to fit our needs throughout 2000/01 and beyond, and that we develop our own facility in working with Voyager, we should expect equilibrium starting in July 2001 -- that is, we should expect to accomplish tasks in Voyager in the same time as it currently takes us to perform the same functions in NOTIS. Thus, one further estimate will apply: FY 2001/02 ... a 0% decrease in current productivity
4. Time required to plan and implement workflow adjustments and for troubleshooting is considered overhead and has not been factored into recovery estimates.
5. CTS functional priorities remain the same for FY 2000/01 and 2001/02.
II. THE SITUATION IN ACQUISITIONS: SECTION-BY-SECTION SUMMARY
Ordering
Priority: High
Downtime: May 16-June 22 (27 business days)
Although in view-only mode, NOTIS will still be available for pre-order searching during the downtime period, as will vendor databases. We can place domestic firm orders for monographs in the Yankee (YBP) database and load the records into Voyager post-implementation (we already know how to do this). We should hold all foreign firm orders for monographs, as well as series and multivol orders, unless they’re rush. Rush orders that cannot be done through YBP will need to be done through the Campus Store, amazon.com, or other means; we’ll need to save the paperwork for post-implementation input. We should also hold off on adding locations, if possible.
Serial orders should be held, unless rush. This includes both domestic and foreign titles and requests for samples. Orders for rushes and print counterparts for titles that have been made available to us in electronic format will need to be done, but their number should be very small and we can keep a paper file for post-implementation input.
We can do pre-order searching in OCLC for non-YBP orders and export records to a temporary file to be later downloaded into Voyager, though it is still unclear whether we will be able do the subsequent downloading in batch mode. Assuming we can figure out how to do so, it will be worth our while to accomplish this part of the ordering task during downtime, then complete the process post-implementation.
We already place many documents orders with major vendors by phone, mail, or e-mail and we can continue to do so during downtime. Orders placed with deposit account vendors are either never recorded in NOTIS or are recorded on receipt as book-in-hand inputting. Books ordered by phone and charged to our procurement card turn into book-in-hand inputting on receipt. Thus, the ordering process in Documents will be only negligibly affected during downtime.
Recovery:
We will need to load the file of domestic firm orders we placed through YBP and eliminate the backlog of monograph and serial orders we set aside or did manually during downtime. We will need to download the file of OCLC pre-order searching into Voyager and complete the ordering process for those titles.
In addition, we will need to clean up our share of some 28,000 open monograph orders system-wide. “Before any activity can occur against these open orders, each one will require system approval and have to be activated to ensure a claiming function” (Scott Wicks, “Shelf-ready Books to Help with Processing Backlog Post Voyager Implementation”). It would be best to hold off for several weeks before updating these open orders en masse. We can process those that match incoming receipts as the items arrive. After a period of, say, 3 months, we can then mount a project to update the remaining orders. This is a fairly high priority in spite of the project’s size and we should aim to complete it within our first year on Voyager.
Numbers:
Special Voyager factors for ordering (see also Assumption #2 in Section I):
Months 4-12 ... a 10% decrease in current productivity
Months 1-4 (composite) ... a 15% decrease in current productivity
FY 2000/01 (composite) ... a 12% decrease in current productivity
Domestic firm orders -- There will be approximately 1500 of these to be done in the YBP database during downtime. We should be able to load them into Voyager post-implementation during the receiving process. Time required for the load should be negligible, once we know how to do it, but we will need to cleanup the order records in Voyager upon transfer. At one minute per title, cleanup will require 25 real time B-level hours.
Manual and foreign orders -- These will need to be input into Voyager. Assuming we will have done pre-order searching in OCLC prior to June 8 and can export those records into Voyager thereafter (i.e. when the cataloging client is available), and that we’ve created the bibliographic portion for order records in Voyager between June 8 and June 22, we can finish the ordering process post-implementation. Estimated time: 2 B-level staff members can accomplish the post-implementation work in 3 days, or 48 real time B-level FTE hours.
Serial orders -- We will be holding up to 40 serial orders during the 27 days of downtime. At 15 minutes per title, we’ll need 10 real time C/D-level hours to catch up post-implementation.
Open monograph orders -- Assuming CTS’s share of these 28,000 orders is 80%, then we’re looking at 22,400 orders to clean up. At one minute per title, that’s 374 real time B-level FTE hours.
Downtime -- Probably no freed up time whatsoever for monograph ordering staff. Although we won’t be placing some 1100 orders, learning new procedures for searching in the YBP and OCLC databases, and downloading OCLC records into Voyager post-June 8 (in addition to Voyager training and practice) are likely to take up most, if not all of that time. For serials staff, we’re looking at roughly 10 freed up real time C/D-level hours, which will easily be taken up by Voyager training and practice.
Recovery -- For new monograph firm order cleanup, 73 real time B-level FTE hours x 1.62 (comp. factor) x 1.25 (Voyager factor) = 148 adjusted B-level FTE hours. For serial orders, 10 real time C/D-level FTE hours x 1.62 (comp. factor) x 1.25 (Voyager factor) = 21 adjusted C/D-level hours. For open monograph order cleanup, 374 real time B-level FTE hours x 1.62 (comp. factor) x 1.12 (Voyager factor) = 679 adjusted B-level FTE hours, assuming clean up is accomplished prior to July 1, 2001. Total = 827 adjusted B-level FTE hours + 21 adjusted C/D-level hours.
Note -- Increased staffing needs just to keep pace with regular work in the Voyager environment for FY 2000/01 will be addressed in Part IV below.
Receiving (excluding APPR items)
Priority: High
Downtime: June 13-22 (invoice payments and claiming -- 8 business days); May 16-June 22 (other tasks associated with receiving -- 27 business days)
We will be able to pay invoices using NOTIS O/P/Rs through June 12, but we won’t be able to update NOTIS copy holdings to complete the receiving process for the duration of the downtime. After June 12, we should pay only those invoices involving a time-sensitive discount or other payment emergency. For those transactions, we’ll need to keep a paper file.
Firm order monographs will need to be held on trucks in the sub-basement for post-implementation processing. Storage information can be recorded in NOTIS copy holdings and printed out, using doubledumb barcodes to simplify subsequent item record creation (one on the item, one on the worksheet). Rush items will need to be processed using annotated printouts and doubledumb barcodes.
We will not be able to complete receipt or check-in of serials for the duration of the downtime, but this material will have to be processed nonetheless. From May 17-June 7, serials receiving staff can check in new receipts in either NOTIS or the Voyager test file, make printouts of the pertinent screens, and update the live Voyager records later. After June 7 (when the Voyager cataloging module is available), staff can check in serials on the Voyager holdings record, adding data from printouts. For hardcover items, we can utilize doubledumb barcodes to simplify subsequent item record creation.
Unless rush, pre-catalog processing of new networked electronic resources should be held until June 8. If an O/P/R is involved, a copy of the electronic resource form should be retained in Acquisitions until normal workflow recommences June 23.
Claiming will not be possible for at least the last 8 business days of the downtime (June 13-22).
Recovery:
We will need to eliminate quickly the backlog of invoices, then turn our attention to the rest of the “unreceived” monographs. These latter items will need to be received, then fastcatted or sent on to Cataloging or storage. Serial check-in and e-resource work will need to be completed in Voyager.
In addition, we will need to mount a project system-wide to create some 53,000 serial check-in records. “Only after these records have been created can we fully utilize the predictive check-in and claiming features of Voyager” (Scott Wicks, “Shelf-ready books ...”). Although we will obviously have to treat this work as an ongoing project, it should be considered a fairly high priority and we should aim to complete it within our first year on Voyager.
Numbers:
Special Voyager factors for firm order receiving (see also Assumption #2 in Section I):
Months 4-12 ... a 10% decrease in current productivity
Months 1-4 (composite) ... a 15% decrease in current productivity
FY 2000/01 (composite) ... a 12% decrease in current productivity
Firm order monographs -- Since we will still be able to pay invoices for the first 19 business days of downtime (i.e. through June 12), we can take care of a portion of the receiving process for some 70% of the 3500 firm order monographs we expect to receive during the period May 16-June 22, or 2450 items. That leaves 1050 titles which will require the full receiving treatment post-implementation. If we assume that a typical receiver can process 8 titles per hour in real time and that having already paid the invoice will allow her to increase this rate to 16 per hour, then 1050 titles at 8 per hour will require 132 real time B/C-level FTE hours, and 2450 titles at 16 per hour will require 154 real time B/C-level FTE hours. Total estimate: 286 real time B/C-level FTE hours.
Networked electronic resources -- Ed estimates that recovery will take “a couple of days.” Let’s figure 10 real time E-level FTE hours.
Serials check-in -- Although we’ll probably want to move them to the front of the queue, serial check-in for issues received during downtime can be subsumed in the numbers for the larger project to create serial check-in records for all open orders campus-wide. Assuming CTS’s share of these 53,000 titles is 80%, then we’re looking at 42,400 records to create. At ten minutes per title, that’s 7067 real time B/C-level FTE hours.
Downtime -- Since we won’t be able to carry out most tasks associated with monograph receiving, some 286 real time B/C-level FTE hours will be freed up. However, organizing the backlog of receipts to be held will take roughly as much time as completing the receiving process for these items. Without students or temps during this period, the net freed up time for monograph receiving staff is likely to be negligible. The Electronic Resources Assistant will have a couple of days freed up, but given his involvement in Voyager implementation we can assume that this time will be swallowed up in overhead. Since serials check-in will go on as usual, albeit in a different manner, serials receiving staff should expect no freed up time; in fact, the interim check-in process is likely to be more time-consuming than serials check-in in NOTIS. Thus, we should count on no freed up downtime for receivers.
Recovery -- For firm order monographs requiring invoice payment (and therefore higher priority), 132 real time B/C-level FTE hours x 1.62 (comp. factor) x 1.25 (Voyager factor) = 268 adjusted B/C-level FTE hours. For firm order monographs, the invoices of which have already been paid, 154 real time B/C-level FTE hours x 1.62 (comp. factor) x 1.15 (Voyager factor) = 287 adjusted B/C-level FTE hours, assuming clean up is accomplished within 4 months. For networked electronic resources (another higher priority), 10 real time E-level FTE hours x 1.62 (comp. factor) x 1.25 (Voyager factor) = 21 adjusted E-level FTE hours. For serial check-in record creation, 7067 real time B/C-level FTE hours x 1.62 (comp. factor) x 1.12 (Voyager factor) = 12,822 adjusted B/C-level FTE hours, if we aim to complete this project by July 1, 2001. Totals = 13,377 adjusted B/C-level FTE hours + 21 adjusted E-level FTE hours.
Note -- Increased staffing needs just to keep pace with regular work in the Voyager environment for FY 2000/01 will be addressed in Part IV below.
Inputting (including APPR items)
Priority: Medium High
Downtime: May 16-June 7 (16 business days)
Without access to NOTIS or the Voyager cataloging module, there is not much we can do to process these items during downtime. However, since there is no NOTIS O/P/R or Voyager payment aspect associated with titles acquired through approval plans, LC programs, exchange, or as gifts, we can begin the online acquisition of these items (and completely process those with copy) as soon as the Voyager cataloging module is available.
Recovery:
Although not as high a priority as the ordering and receiving backlogs, these titles will need to be processed with some urgency after June 7, since they will not be represented in the catalog until inputting is complete.
Numbers:
Special Voyager factors for inputting (see also Assumption #2 in Section I):
Months 4-12 ... a 5% decrease in current productivity
Months 1-4 (composite) ... a 14% decrease in current productivity
FY 2000/01 (composite) ... an 8% decrease in current productivity
Based on 1998/99 statistics, we should expect to receive 1800 monographic titles through approval plans, LC programs, etc. during downtime. At 12 titles per real time hour, we will be able to eliminate this backlog in 150 B-level FTE hours.
Downtime -- 150 real time B-level FTE hours will be freed up. However, since there are many staff members involved in this task, these hours, spread out among so many, will probably not add up to much, especially taking into consideration Voyager training and practice.
Recovery -- 150 real-time B-level FTE hours x 1.62 (comp. factor) x 1.14 (Voyager factor) = 277 adjusted B-level FTE hours, assuming we catch up within 4 months.
Note -- Increased staffing needs just to keep pace with regular work in the Voyager environment for FY 2000/01 will be addressed in Part IV below.
Physical Processing
Priority: Medium (High for rush items)
Downtime: None (except for most automated labeling: May 16-June 7, 16 business days)
Although physical processing can continue without NOTIS or Voyager, we’ll need to create spine labels manually, when needed, during downtime. We will be able to plate items being held on receipt trucks (see above under “Receiving”) and complete physical processing for rush items. Ongoing hardcover serials and hardcover or stiffened added volumes can be completely processed and labels generated automatically through NOTIS, even during downtime, since there will already be a call number in the copy holdings record for use on the spine label (though this approach will also require serials check-in staff to create item records in NOTIS and utilize doubledumb barcodes -- one on the item, the other on the check-in worksheet).
Recovery:
Although we do not expect a backlog of physical processing per se, we will eventually have to deal with a bubble of material resulting from ramped up efforts to eliminate the post-implementation receiving, inputting, and cataloging backlogs.
Numbers:
Assumptions -- We average 20 items per hour for complete physical processing, 30 items per hour for plating only, and 30 items per hour for marking only. We do not expect these averages to change with Voyager, so no Voyager factor applies to physical processing.
Downtime -- There will be 3500 firm order receipts, all of which can be plated, as well as a certain number of hardcover serials and added volumes which can be completely processed. Cataloging should be able to send through nearly 500 new titles, most of which will require marking only (since they’ll be coming from the backlog). That’s a total of 4000+ items, a significant drop from the 12,840 titles (chiefly processed at the 30/hr. rate) that we averaged during a typical 3.4 week period in FY 1998/99. Even with Voyager training and practice, we’re probably looking at 50% of all A-level staff time freed up during downtime, and perhaps 25% for the two weeks or so immediately thereafter.
Recovery -- This is difficult to predict, because it will depend a great deal on what overall strategy we adopt throughout the department. Assuming, however, that we aim to achieve a steady state working backlog by the end of FY 2000/01, we will need to catalog some 98,000 new titles during the period May 16, 2000-June 30, 2001 (based on 1998/99 production and backlog reduction numbers and recovery estimates from elsewhere in this report). That’s a drop of nearly 40,000 titles from our 1998/99 pace (i.e. 137,924 new titles for 58 weeks), or 28%, suggesting that we will be able to free up roughly 1.0 FTE out of the 3.5 regular FTE currently assigned to Physical Processing to work on other post-implementation tasks. If we allow small, short-term cataloging backlogs to develop while we complete cleanup in Acquisitions (see Sections IV and V below), we will be able to free up slightly more staff time in Physical Processing during FY 2000/01.
III. THE SITUATION IN CATALOGING: SECTION-BY-SECTION SUMMARY
Backlog Project
Priority: High, but will be completed before May 16.
Copy Cataloging
Priority: Medium (High for rush items)
Downtime: May 16-June 7 (16 business days)
Since Receiving and Inputting will be pretty much out of commission from May 16 to June 22, there will be no new receipts trucks to attend to. We will still need to process rush items, but these can be done using CatME (or, in a pinch, print worksheets) and doubledumb barcodes. We can also continue to process straightforward added volumes, using worksheets and doubledumb barcodes.
We can continue to search and catalog items from areas of the backlog unaffected by pending MARCadia runs using CatME, storing updated records in “export.dat” files, and downloading these records into Voyager post-implementation. This might be a particularly good time to work on the microforms backlog. CJK processing will be only minimally affected, since that work is done in RLIN prior to downloading into the local catalog. We will simply have to delay the downloading aspect of the process until after June 7.
Recovery:
We will need to download “export.dat” files into Voyager and transfer written information from worksheets into online records. Once the backlog of receiving and inputting starts coming through to Cataloging, we will need to ramp up work on new receipts.
Numbers:
Assumptions -- A cataloger can process 12 MARCadia items per hour (real time), 4.5 non-fastcat copy items per hour (real time), 3 MLC items per hour (real time), and 1.5 original items per hour (real time).
Special Voyager factors for copy cataloging (see also Assumption #2 in Section I):
Months 4-12 ... a 5% decrease in current productivity
Months 1-4 (composite) ... a 14% decrease in current productivity
FY 2000/01 (composite) ... an 8% decrease in current productivity
New receipts -- We receive on average 6 titles per day on the Copy Cataloging new receipts truck. Consequently, during the 27 days of Receiving downtime, approximately 162 items will have been held. At 4.5 titles per hour, 36 real time D-level FTE hours will be required to eliminate this relatively small backlog of priority items.
Downtime -- If we assume that roughly 50% of cataloging downtime can be devoted to work on the backlog and that searching and cataloging on OCLC will take roughly twice as long as work on NOTIS (given the learning curve and this short 16-day period), the Copy Cataloging Unit will be able to produce roughly ¼ of its normal output during downtime. The total output of the Copy Cataloging Unit during FY 1998/99, not including the work of backlog project students and temps, was 17,900 titles, or 344 titles per week (based on a 52-week year). Thus, during the 3.4 weeks of downtime (including the holiday), we would normally expect an output of 1169 titles. 25% of 1169 = 292 titles cataloged during downtime. No freed up time.
Recovery -- Although we will be faced with a 2-4 month hiatus in our MARCadia runs as a result of Voyager implementation and the time it will take to rewrite the program for use with Voyager, and although we might need to face a mini-blobzilla when we restart the program, we should assume continued use of both MARCadia and minimal level cataloging to cope with post-implementation recovery. Since we expect this unit to catalog only 25% of what it normally would during the downtime period, that difference is 1169 minus 292, or a shortfall of 877 titles. In the 55 weeks following the implementation of the Voyager cataloging module, we should be able to process 344 x 55 = 18,920, divided by 1.08 (Voyager factor) = 17,518 titles: a net loss of 1402 titles from normal production pace. Total copy cataloging shortfall through July 1, 2000: 2279 titles.
Original Cataloging
Priority: Medium (High for rush items)
Downtime: May 16-June 7 (16 business days)
Original Cataloging’s situation is similar to that of Copy Cataloging. Since Receiving and Inputting will be pretty much out of commission from May 16 to June 22, there will be no new receipts to review. We will still need to process rush items, but these can be done using CatME (or, in a pinch, print worksheets) and doubledumb barcodes.
We can search and catalog items from areas of the backlog unaffected by pending MARCadia runs using CatME, storing updated records in “export.dat” files, and downloading these records into Voyager post-implementation. New serials can wait until we’re up on Voyager. We will still be able to catalog electronic resources on the Gateway, but since creating the Gateway record prior to completing the cataloging for the OPAC is inefficient, it would be best to hold this work until Voyager and the Voyager/Gateway interface are ready -- except for rush titles, of course. CJK processing will be only minimally affected, since that work is done in RLIN prior to downloading into the local catalog. We will simply have to delay the downloading aspect of this process until after June 7.
Recovery:
We will need to download “export.dat” files into Voyager and transfer written information from worksheets into online records. Once the backlog of receiving and inputting begins coming through to the review shelves, we will have more titles to review (though not necessarily more titles to catalog -- this ratio may vary from area to area). We should also be prepared to ramp up work on serials and other priority items (e.g. first volumes in multivol sets) once the receiving and inputting bubble makes it way to Cataloging.
Numbers:
Assumptions -- A cataloger can process 12 MARCadia items per hour (real time), 4.5 non-fastcat copy items per hour (real time), 3 MLC items per hour (real time), and 1.5 original items per hour (real time).
Special Voyager factors for copy cataloging (see also Assumption #2 in Section I):
Months 4-12 ... a 5% decrease in current productivity
Months 1-4 (composite) ... a 14% decrease in current productivity
FY 2000/01 (composite) ... an 8% decrease in current productivity
Downtime -- If we assume that roughly 50% of cataloger downtime can be devoted to work on the backlog and that searching and cataloging on OCLC will take roughly twice as long as work on NOTIS (given the learning curve and this short 16-day period), the Original Cataloging Unit will be able to produce roughly ¼ of its normal output during downtime. The total output of the Original Cataloging Unit during FY 1998/99 was 11,749 titles, or 225 titles per week (based on a 52-week year). Thus, during the 3.4 weeks of downtime (including the holiday), we would normally expect an output of 765 titles. 25% of 765 = 191 titles cataloged during downtime. No freed up time.
Recovery -- Since we expect this unit to catalog only 25% of what it normally would during the downtime period, that difference is 765 minus 191, or a shortfall of 574 titles. In the 55 weeks following the implementation on the Voyager cataloging module, we should be able to process 225 x 55 = 12,375, divided by 1.08 (Voyager factor) = 11,458 titles; a net loss of 917 titles from normal production pace. Total original cataloging shortfall through July 1, 2000 = 1491 titles.
Recon (P/D/E)
Priority: High, but will be completed before May 16.
IV. FURTHER DISCUSSION AND ANALYSIS: THE BIGGER PICTURE
Acquisitions
Although increased staffing needs to eliminate downtime backlogs were addressed in Section II, we will also need to keep in mind that pre-implementation levels of staffing in Acquisitions will be insufficient to keep pace with regular work during the first year of Voyager -- at least for ordering, receiving, and inputting (although not for physical processing). We currently have 31.5 FTE in these three areas, excluding managers.
To keep pace with regular work (i.e. not counting the recovery work outlined in Section II), we would need 31.5 x 1.11 (a cross-functional Voyager factor, excluding physical processing) = 34.96 FTE for FY 2000/01. If we free up 1.0 FTE in Physical Processing because of the expected decline in output, that reduces the total need to 33.96. If we implement a shelf-ready books program (see “Other options” below), we can further reduce our total need to 33.59 -- overall, an increase of 2.09 FTE for FY 2000/01.
The student budget for FY 1999/2000 calls for 7920 student hours. This is likely to remain the same for FY 2000/01. Although not all student work in Acquisitions is Voyager-related, let’s hedge and apply a cross-functional Voyager factor of 1.08 (includes all ACQ functions) to the entire budget. 7920 x 1.08 = 8554 hours. That is, for the first post-implementation year, we would need an additional 634 student hours to keep pace.
Now, let’s recall the recovery pieces from Section II:
Ordering: 827 adjusted B-level FTE hours; 21 adjusted C/D-level FTE hours
Receiving: 13,377 adjusted B/C level FTE hours; 21 adjusted E-level FTE hours
Inputting: 277 adjusted B-level FTE hours
Phys Proc: does not apply
Total: 14,523 hours (B-level and higher)
An adjusted FTE year is 39 hrs. x 52 wks. = 2028 hours. Thus, the 14,523 hours of recovery time outlined in Section II represents the work of 7.16 FTE for one year. Adding in the 2.09 FTE to keep pace with regular work for the first year (let’s leave the student hours aside for the moment), we get roughly 9.25 additional FTE required to recover in Acquisitions by the end of FY 2000/01. Given the overall high priority of the work involved, we should not aim to prolong this aspect of recovery beyond June 30, 2001.
Cataloging
The situation in Cataloging is more promising, assuming that the backlog project and P/D/E recon are indeed complete by May 16.
Using the 1998/99 production and backlog count numbers, we note that 123,660 new titles were cataloged during that single year and the backlog was reduced by 36,380 pieces. Subtracting the latter from the former number, we get 87,280 new titles received. Since the backlog count is a pieces count, rather than a title count, let’s round up that number to 88,000. Now, let’s be conservative and assume that all Copy Cataloging students and temps were paid out of backlog project money (which is no longer in the budget). This latter group cataloged 21,790 items. So without their production (but utilizing MARCadia and MLC), we would have still processed 123,660 minus 21,790 = 101,870 titles with regular fastcat and cataloging staff (whose FTE numbers are still roughly the same, though slightly redistributed). That’s an estimated 13,870 more titles cataloged than received.
Thus, all else being equal, we might expect a production surplus of 13,870 titles now that the backlog project is complete. However, from the cataloging shortfalls outlined in Section III (fastcat shortfalls in Acquisitions are covered in recovery needs for Receiving and Inputting) , we know that we’ll have to catalog an additional 2279 + 1491 titles = 3770 to make up for downtime. Applying the Voyager factor, we get 3770 x 1.08 = 4072, meaning that it’ll take us as much time to process those 3770 backlogged titles as it normally would for 4072 titles. That reduces our potential production surplus for FY 2000/01 to 13,870 minus 4072 = 9798 titles.
In FY 1998/99, the Copy Cataloging Unit processed 17,900 new titles (not counting student and temp contributions) and the Original Cataloging Unit processed 44,505 new titles for a total of 44,505 new titles processed. Applying the Voyager factor for FY 2000/01, that’s 29,649 divided by 1.08 = 41,208 titles, or a decrease of 3297 titles. That reduces our potential production surplus for FY 2000/01 to 6501 titles.
Nonetheless, we still have a surplus. 6501 titles represents 36% of the 17,900 titles the Copy Cataloging Unit produced in FY 1998/99 (excluding student and temp production). Our staffing at the D-level (not counting Laurie Stevens = 5 FTE) will be roughly the same as of July 1, 2000. This means that we can free up approximately 1.8 FTE worth of D-level staff to help with recovery in Acquisitions for one year without allowing a cataloging backlog to grow. This would assume some reapportionment of cataloger focus, of course, as original catalogers would have to do significantly more copy, but it would save us considerable recovery expense on the departmental level.
We could also project recovery over a longer period, if we are willing to live with a cataloging backlog for that period of time . Let’s assume that in addition to lending 1.8 FTE of D-level staff to the cleanup effort in Acquisitions for the duration of FY 2000/01, we also pledge a certain amount of original cataloger time (that is, actual cataloging time) during FY 2000/01. Assuming that original catalogers (E-level and librarians, combined) spend 75% of their total adjusted time actually cataloging, that’s approximately 15 FTE x .75 = 11.25 FTE of cataloging.
At 1 day per week, that’s 20% of 11.25 FTE or 2.25 FTE of original cataloging staff to lend to Acquisitions cleanup for one year. Since original catalogers processed 26,605 titles during 1998/99, we should expect 26,605 divided by 1.08 = 24,634 for FY 2000/01. Cutting this production by 20% would result in 24,634 x 0.20 = a backlog of 4927 titles by June 30, 2001. However, we can deduct the Voyager factor from this total (since we will not be processing these items until FY 2001/02 when the Voyager factor = 0). 4927 divided by 1.08 = a backlog of 4563 titles. Recalling that the cataloging surplus, without the decreases in production expected for the first 52-55 weeks post-implementation, would otherwise be 13,870 titles per annum, we can project a backlog reduction pace of 1155 titles per month, once we have completed the cleanup in Acquisitions and once we’ve begun our second year on Voyager. Thus, if we use regular Cataloging staff at a moderate level (a total of 4.05 FTE) to help with Acquisitions cleanup during FY 2000/01, we can expect to eliminate the resulting cataloging backlog by Oct. 31, 2001 (4563 divided by 1155 per month = 3.95 months).
If we increase original cataloger participation in Acquisitions cleanup by raising the commitment to 2 days per week, that’s 40% of 11.25 FTE or 4.5 FTE of original cataloging staff to lend to Acquisitions cleanup for one year. Cutting this original cataloger production by 40% for FY 2000/01 would result in 24,634 x 0.40 = 9854 titles. Deducting the Voyager factor, that’s 9854 divided by 1.08 = a backlog of 9125 by June 30, 2001. We can eliminate that backlog in under eight months (9125 divide by 1155 = 7.90 months). Thus, if we use Cataloging staff at a bit more intensive level (6.3 FTE) to help with Acquisitions cleanup during FY 2000/01, we can expect to eliminate the resulting cataloging backlog by Feb. 28, 2002.
If we increase both copy and original cataloger participation to 50%, we would free up 2.5 D-level FTE and 6.75 E-level and academic FTE to assist in Acquisitions cleanup. Cutting the production of our D-level catalogers by 50% for FY 2000/01 would result in a backlog of some 3112 titles by June 30, 2001. Doing the same to original cataloger production over the same period would result in a backlog of some 11,407 titles. 3112 + 11,407 = 14,519 new titles. Deducting the Voyager factor, that’s 14,519 divided by 1.08 = a backlog of 13,444 new titles by June 30, 2001. Thus, if we use Cataloging staff very intensively (9.25 FTE) to help with Acquisitions cleanup during FY 2000/01, we can expect to eliminate the resulting cataloging backlog in under 12 months (13,444 divided by 1155 = 11.63 months), or by June 30, 2002.
The student budget for FY 2000/01 calls for 1125 student hours. Applying the Voyager factor, that’s 1125 divided by 1.08 = 1041 equivalent student hours for FY 2000/01. If we pool Acquisitions and Cataloging students for the post-implementation recovery, the student hours from Cataloging will more than make up for the additional student needs in Acquisitions. Since student productivity was not included in the analysis of cataloging needs in the preceding paragraphs, whatever cataloging they do contribute will be a bonus.
Other options
Shelf-Ready Books from Yankee:
This proposal, outlined in Scott’s “Shelf-ready books to Help with Processing Backlog Post Voyager Implementation,” would free up 0.5 C-level FTE and .25 A-level FTE for six months, or a net total of 0.37 FTE over a whole year at a cost of $19,210. Although this plan would by no means save us money (salary savings would only amount to some $10,000 for this 6-month period), its implementation would be advantageous for two reasons: (1) it would free up trained staff and reduce the number of new staff we would need to train during the busy recovery period, and (2) it would serve as a valuable experiment in further outsourcing. For these reasons, I’ve included it in the four recovery scenarios outlined in Section V below.
Using a Harvard/Hoyt/Hillmann Program to Create Serial Check-In Records:
There’s been some discussion about using ISSNs and Harvard check-in patterns to create a file of records that would significantly reduce the amount of time and level of staff needed for the project to create 53,000 serial check-in records system-wide. Given the huge demand on staff time this project represents, such a solution is surely worth pursuing, but we do not yet know how much staff savings we can expect.
Purchasing an Additional CCD Scanner for Physical Processing:
The price tag on this tool is $450, but it would increase the speed at which marking can be done in the new environment. Again though, it is difficult to quantify potential staff savings at this point.
Downtime
With the exception of Physical Processing staff, who will have a significant amount of time on their hands, most staff downtime will be taken up in restructured routines for accomplishing whatever regular work we can still do and with Voyager training and practice. Nonetheless, there are a couple of other tasks to which we can redirect staff time, if necessary:
For Physical Processing staff, only the first seems appropriate, so we are still in need in finding something else for them to do for a fraction of the 16 business days during which no cataloging will be accomplished.
V. CONCLUSION
Assuming we would hire additional term-appointment staff at the B-level to assist in recovery (or to back up regular staff who are doing higher level recovery work), we end up with the following, based on the cross-unit analysis in Section IV:
To recover in one year (i.e., by June 30, 2001) --
Acquisitions Needs 9.25 FTE
Cataloging Can Lend 1.8 FTE
Difference 7.45 FTE
Costs: Salary - $26,325 x 7.45 = $196,120; Shelf-ready books - $19,210; CDC scanner - $450
Total Cost $215,780
To recover in 16 months (i.e., by Oct. 31, 2001) --
Acquisitions Needs 9.25 FTE (thru June 30, 2001)
Cataloging Can Lend 4.05 FTE
Difference 5.20 FTE (thru June 30, 2001)
Cost: Salary - $26,325 x 5.20 = $136,890; Shelf-ready books - $19,210; CDC scanner - $450
Total Cost $156,550
To recover in 20 months (i.e., by Feb. 28, 2002) --
Acquisitions Needs 9.25 FTE (thru June 30, 2001)
Cataloging Can Lend 6.30 FTE
Difference 2.95 FTE (thru June 30, 2001)
Cost: Salary - $26,325 x 2.95 = $77,660; Shelf-ready books - $19,210; CDC scanner - $450
Total Cost $97,320
To recover in 24 months (i.e., by June 30, 2002)
[Note: in earlier version, 18 months, or by Dec. 31, 2001]
Acquisitions Needs 9.25 FTE (thru June 30, 2001)
Cataloging Can Lend 9.25 FTE
Difference None
Cost: Salary - $0; Shelf-ready books - $19,210; CDC scanner - $450
Total Cost $19,660
The foregoing analysis suggests that recovery from Voyager implementation may not be as grim as the LMS Evaluation Committee predicted almost a year ago, though it will not be without its challenges. As I remarked in Section I, some of the data on which I’ve based my calculations is quite impressionistic. Moreover, assumptions about how long it’s going to take us to do our work in Voyager 99, or in Voyager 2000, are mere shots in the dark. Finally, the assumption that new receipts have remained at a steady state for past year and will continue to do so through the next 15, 18, 23, or 27 months is obviously shaky (although we’ll know more after we’ve evaluated this year’s backlog count).
I’ve played it safe with some on my calculations for our potential use of student hours, and I’ve rounded figures conservatively in many cases. Nonetheless, it might still be wise to hedge a bit more in our estimates of either cleanup time or cleanup cost.
Submitted by: Jim LeBlanc, 3/31/00
Revised numbers to address error in 1998/99 OC production figure, 4/11/00
