Pick a collection section the team can finish, pull its circulation data and give the flagged titles a closer look. Before the first cart leaves the workroom, write down how accuracy, condition, demand, access to another copy and community needs will shape the decisions. That gives the selector a clear task and leaves the next reviewer a record they can use.
1. Pull the Right Circulation Report Before Touching a Shelf
Start with your integrated library system's item-level reports or ask the person who manages reporting for an extract. Check what each field means in your system before relying on a canned “weeding list.” Useful fields may include barcode, title, author, call number, shelving location, collection code, status, item-added date, checkouts in the chosen period, last checkout date, and system-wide copy count. Include in-house use if it is recorded consistently. Price can inform a replacement discussion, but the amount paid for an item is not a measure of its present usefulness. Ask for item statistics without patron names or borrowing histories; ALA’s privacy guidance[1] calls for reports to contain only the personal data needed for operations.
Choose a reporting period that makes sense for each collection. A recently added item has had less chance to circulate than an older copy; a specialized reference work may be consulted without a checkout; seasonal demand can disappear in a short window. Record the dates, eligible statuses, branch scope, and whether renewals count as checkouts. Export a working copy to a staff location with appropriate access controls, then spot-check several known items against their catalog records. Fix missing statuses, transfers, or misleading counts before treating the spreadsheet as a shelf list. A clean report saves repeated work, but it remains a screening tool.
2. Set Cutoff Rules That Match Your Building and Budget
Before anyone scans a barcode, read the library's current collection-development and deselection policy. Record which collections may be reviewed, who makes final decisions, and what criteria apply. A local checkout threshold can create a manageable candidate list if the policy allows it. Document the period and collection to which that threshold applies, and treat every flag as a prompt for examination. Publication date, accuracy, physical condition, format, demand, and access to other copies all matter. Two staff members should be able to see why an item was flagged, while a trained selector retains room to explain an exception.
Build local screening thresholds, then test them
There is no single checkout count that fits adult fiction, children's books, reference works, and media. Start with a small sample from one collection: sort by circulation, pull a few low-use and recently acquired items, and compare the list with what is on the shelves. Are the items outdated, damaged, hard to find, temporarily misplaced, or simply serving a small audience? Review the sample with a selector familiar with that subject and branch. Only then choose a provisional date range or checkout count for that collection and note why it is appropriate. Test the rule again before applying it across branches. An infrequently used local-history title or the consortium's only lendable copy may merit retention even when a popular-fiction copy with the same checkout count does not. Keep audiobooks and other formats in their own comparison groups rather than assigning them a universal cutoff.
Budget-linked exceptions
With a smaller materials budget, check replacement options early so a necessary withdrawal does not leave an avoidable gap. Before withdrawing a low-use title, ask whether readers could still obtain an accurate, accessible copy through this library, another branch, or a loan agreement that actually serves them. Check the cost and availability of replacement where those matter. Space pressure may justify a priority review, but it does not prove that the scarcest titles are expendable. Record the trade-off in the same policy-based process used when the budget is healthy.
A costly art book, a complete multi-volume set, or a title no neighboring library will lend deserves closer attention. The original purchase price alone cannot justify an extra year on the shelf, and a cheaper item is not automatically replaceable. Compare current demand, condition, relevance, and the practical availability of an equivalent item. If removing one volume makes a set less usable, review the set together. Flag these cases before a large pull and route them to the staff member responsible for the collection instead of encoding an arbitrary price floor into the report.
3. Clean the Report Before It Becomes a Cart List
A low count can have more than one explanation. Check current holds, loans, transfers, repair and processing statuses before printing a shelf list. An active hold calls for a service decision before withdrawal; it is not a permanent exemption. If a replacement is on order, verify what will remain accessible while the new copy is pending. A matching ISBN does not guarantee that a replacement has arrived, been cataloged, or is ready to circulate. Keep those items on a pending list so the report does not turn an anticipated replacement into a gap on the shelf.
Investigate apparent duplicate barcodes or records with technical services before acting. A shared system may show the same title in several branches or formats, and two report rows do not necessarily represent two physical copies. Missing or long-unavailable items need their own inventory and catalog workflow; they are not books a shelf worker can pull. Mark status and copy-count uncertainties for resolution, then rerun or annotate the candidate list. Keep the difference between withdrawing a present item and cleaning an inaccurate catalog record visible in the final log.
Floating collections and shared branches make location especially important. An item that is quiet in one building may have served readers elsewhere; a title held by one branch may be the only borrowable copy in the system. Compare item and title records, current holdings, requests, and loan options before judging the copy. When demand is concentrated at another location, a transfer or better display may answer the problem. When every copy is worn or outdated, the collection may need replacement rather than simply fewer items. Ask the selector for that subject or consortium to review uncertain cases.
4. Manage the Edge Cases Without Letting Them Stall the Project
Give local authors, scarce editions, and specialized reference works a review route with an assigned decision-maker. Record why each item needs closer attention and set a realistic date under staffing and policy limits. That keeps difficult cases moving without letting a numerical cap send an unexamined title into the discard stream.
Quick review pile
A useful review pile has enough information to resolve each case: item identifier, collection, usage period, condition, other holdings, known requests, and the reason it was flagged. The selector can retain, repair, replace, transfer, or withdraw it under the library's policy. If a title stays, note the reason in the project log so the next cycle can reconsider it with context. Assign the decision to the person authorized by local policy, with enough review time to settle the recorded questions. Escalate disputed or controversial titles through the regular policy rather than using weeding as a shortcut around reconsideration.
Local history and reference
For local-history and reference material, check in-house consultation, scarcity and preservation purpose before interpreting a low checkout count. Follow the collection's own policy and ask the librarian responsible for it to review condition, duplicates, provenance, rights, and access. Do not discard an original merely because a digital image exists: confirm that the digital version is complete, usable, durable, and appropriate for the collection's obligations. For reference works, check whether information remains accurate and whether a reliable replacement is genuinely available. In-house use data can help if captured consistently, but an empty count or a staff member's lack of recollection is weak evidence of no use. ALA notes[2] that libraries may exempt genealogy and local-history collections from ordinary weeding rules.
5. Turn the Clean List into a Staff Workflow
Break the candidate list into shelf-location sections and assign each to staff who have the time and training to inspect items. Set workloads and completion dates from actual staffing, shelving density, and the library's review process. One small section may fit a shift; a rare or damaged collection may take longer. Give the worker the report date and clear instructions for recording exceptions. The aim is a controlled handoff from data review to physical inspection, not a target number of barcodes scanned per hour.
At the shelf, verify the barcode and item record, then check condition, edition, and whether the title is where the report says it is. Record an item as “under review” only if the ILS and local procedure support that status. Keep carts and work lists identifiable so items are not lost to patrons or staff. A damaged copy may call for repair or replacement; a sound but overlooked copy may need a display or transfer. Neither outcome follows automatically from a low checkout count. Have an authorized selector make the final removal decision and follow the library's catalog and disposal procedure. Choose a project schedule that lets staff finish each section without leaving a long-lived, unexplained cart in the workroom.
6. Record What You Pulled and Check the Results Later
Record the reporting period, screening rules, decisions, and disposition counts by collection. Keep a short reason for exceptions and for withdrawn items with unusual access or preservation implications. Store the log where the responsible team can find it for the next review, under the library's data-retention and access rules. This record makes the process understandable to staff and leadership without preserving patron-level borrowing information.
After a suitable interval, compare the reviewed section with its earlier baseline. Look at total loans, holds or requests, collection size, available copies, item age, and gaps readers reported. If useful, compare the same season in a prior year, while noting changes in hours, displays, acquisitions, and catalog exposure. An increase in loans per remaining item can occur simply because the number of items fell; it does not by itself show that a cutoff was sound. Check whether readers still have access to needed titles, including low-demand subjects and formats, before adopting the same threshold elsewhere.
A concise report to staff or the board can show how many items were retained, repaired, transferred, replaced, and withdrawn, together with the reasons and any service gaps found. Describe circulation changes as observations, not proof that weeding caused them. If requests rise for something removed, revisit that decision. If a section remains hard to use, inspect discoverability, shelving, accuracy, and acquisition needs before tightening the cutoff again. The next cycle should be informed by both the numbers and what staff and patrons learned.
Use circulation data to direct staff attention, then record the policy and professional reasoning behind each decision. Leave time for difficult titles, protect access and patron privacy, and keep the results available for the next cycle. The review becomes repeatable because staff can follow the evidence and the exceptions, not because every item met the same numerical cutoff.