Choosing an LMS with Two Staff and a Four-and-a-Half Lakh Grant

The licence price is not the cost. Compare three-year cash outlay:

• Vendor A: 3.8 lakh licence + 1.2 lakh migration + 40,000 x 3 maintenance = 6.2 lakh (before any hardware and barcode scanners).
• Vendor B: 2.9 lakh implementation/migration/support year one + 35,000 x 2 support = 3.6 lakh (assuming the quote includes the same scope).

Koha is cheaper and is open source, but 'free software' is not 'free implementation'. The decisive questions are local support, Unicode Marathi, and migration quality. Recommend Vendor B with written conditions:

1. A sample migration of 500 records, including Marathi titles, accepted before full payment.
2. A written Unicode test for Marathi input, search, and display in the OPAC.
3. Training for both clerical staff and one backup person, with a manual in Marathi.
4. A support SLA with response times and a named local engineer.
5. An exit clause and a full data export in MARC or CSV at any time.

If the partner cannot demonstrate Marathi and migration, Vendor A is the safer fallback even at higher cost. The recommendation is conditional, not ideological.

Do not attempt a big-bang migration. Sequence the work:

• Month 1: clean the 4,000 Excel records; define minimum fields (author, title, publisher, year, call number, accession number, language). Buy two barcode scanners and labels; assign accession numbers to unnumbered books.
• Month 2: migrate the cleaned Excel set and test retrieval; do not load the 8,000 unretroconverted books yet.
• Month 3: pilot barcoding and circulation on 500 high-circulation books; train staff on issue, return, and renewal.
• Month 4: outsource or batch-create short MARC records for the 8,000 books, using shelf-list cards and the accession register. Accept short records with author, title, call number, and subject if full cataloguing is impossible.
• Month 5: load the short records; run duplicate checks; configure the OPAC in Marathi and English; test diacritics and search.
• Month 6: public launch, user orientation, and a printed guide.

Staff capacity is the binding constraint: two clericals cannot both migrate and run the counter. Ring-fence two half-days per week for data work and defer non-essential activities. Measure progress by records loaded and barcodes applied, not by vendor demonstrations.

Fear is a management problem, not a training problem. Address it directly and early.

• State the staffing decision in writing: the system will not reduce sanctioned posts; it will change tasks. If redeployment is possible, say so before rumours harden.
• Redesign roles: one staff member becomes the circulation and barcode superuser; the other becomes the data-quality and OPAC superuser. Give each a named responsibility and backup.
• Train on the actual work, not generic software: issuing a book, finding a Marathi title, renewing a member, and generating an overdue list. Use the library's own records.
• Involve staff in decisions such as label placement and counter workflow; ownership reduces resistance.
• Protect against blame: agree that migration errors will be logged and fixed centrally, not blamed on the counter staff.
• Measure and recognise: report time saved per transaction and errors reduced; acknowledge the staff publicly.

If resistance persists, use a written performance discussion about duties, not a threat. The governing body should also hear that the system requires at least one trained backup; otherwise the library becomes dependent on one person.