The $17,000 Tax You Don't Have To Pay

Alright. You've read Parts 1 and 2. You understand what happened with Baker & Taylor. You know the contract traps. You're pissed off about paying $0.59 for MARC records that come from free public databases.

Now let's do something about it.

This post is going to show you exactly how to do your own cataloging and processing. I'm writing it for people who have never opened a command line, never written code, and break into a cold sweat when someone says "API."

If that's you, stay with me. This is doable.

What You're Actually Paying For

When you pay a vendor for "cataloging and processing," you're paying for four things:

1. A MARC record - The standardized bibliographic data that goes into your ILS
2. Labels - Spine labels, barcode labels, ownership stamps
3. Physical processing - Mylar jackets, lamination, security strips
4. The labor to put it all together

Here's what that costs from a typical vendor:

(Note: This is for standard hardcover processing. Paperback lamination adds $2.15-2.50 per item.)

For a library processing 5,000 items per year, that's $18,100-23,100 annually.

Now here's what it actually costs to do it yourself:

For that same 5,000 items: $3,750-5,750 annually.

You could save $12,000-17,000 per year.

"But what about staff time?" I hear you. This isn't free labor - someone has to do the work. But here's the math: at $25/hour, you'd need to spend 480-680 hours on processing to break even with vendor costs. That's one part-time position doing nothing but processing all year. For most libraries, the actual time investment is far less - especially once you've set up your workflow.

Part A: Getting MARC Records (The Easy Way)

The Library of Congress maintains a free, public database of MARC records. They have records for millions of books. You can search it right now, for free, without creating an account.

Option 1: The Web Interface

1. Go to: catalog.loc.gov
2. Click "Search"
3. Enter your ISBN in the search box
4. Find your book in the results
5. Click on the title
6. Look for "Download" or "Export" options
7. Choose "MARC (UTF-8)" format
8. Save the file

That's it. You just got a MARC record for free.

Option 2: Ask AI To Do It

Open Claude (or whatever AI assistant you use) and say:

"I need to look up the MARC record for ISBN 978-0316769488 from the Library of Congress. Can you query their catalog and give me the bibliographic information?"

Claude can search the Library of Congress catalog and return the data you need - title, author, publisher, publication date, subjects, call number.

Option 3: Batch Processing

For higher volume, tell Claude:

"I need a simple tool that takes a list of ISBNs and looks up each one in the Library of Congress. For each book, I want: title, author, publisher, publication date, and call number. Give me the results in a spreadsheet I can download."

Claude will build this for you. Right there in the conversation. No coding required.

Part B: Generating Labels

Labels are just PDFs with text and barcodes positioned in the right places. Your vendor makes this seem complicated. It isn't.

Here's the prompt:

"I need to generate processing labels for library books. For each book, I need a spine label with the call number on 3 lines, a barcode label with Code 39 barcode and the number printed below, and an ownership label with our library name. Our library is [Your Library Name]. Format it to print on standard Avery 5160 label sheets."

Claude will generate an HTML file that you can open in your browser and print.

Matching Your Current Format

If you want labels that match exactly what your vendor used to send, you'll need to measure your old label sheets and tell Claude the exact dimensions. You can also get custom die-cut label stock made (one-time cost, maybe $500-1500 for the die).

Or you can just switch to a standard format like Avery labels. Your staff will adjust in about a day.

Part C: Getting Label Stock

Standard Label Stock: Avery and other brands make label sheets in standard sizes. You can buy these at any office supply store. Common formats include Avery 5160 (30 labels per sheet), Avery 5163 (10 labels per sheet), and Avery 5167 (80 labels per sheet).

Custom Die-Cut Stock: If you want to match your old vendor's format exactly, take one of their label sheets to a local print shop. They can create custom label stock that matches those exact positions.

The Complete Workflow

Here's what your new process looks like:

1. Receive books - Books arrive. Pull them out of the box.
2. Gather ISBNs - Scan or type the ISBNs into a spreadsheet.
3. Get MARC records - Upload your ISBN list to Claude. Download the MARC file. Import into your ILS.
4. Generate labels - Same spreadsheet, add your barcode numbers. Upload to Claude. Download the PDF. Print on your label stock.
5. Apply labels - Peel and stick.
6. Physical processing - Apply mylar jackets, security strips, etc.

"What About Quality Control?"

Fair question. Vendors have (theoretically) quality-checked their MARC records. What if the Library of Congress record has errors?

Here's the thing: vendor records have errors too. Anyone who's done cataloging knows you're always cleaning up vendor records. Wrong subjects, missing fields, inconsistent formatting.

The Library of Congress records are generally high quality - they're created by professional catalogers following national standards. For most trade books, LC records are as good or better than what vendors provide.

"This Still Sounds Hard"

Let me be honest about the learning curve.

What's actually easy:

• Looking up records one at a time via the LC web interface
• Using Claude to look up ISBNs conversationally
• Generating labels for a few books at a time

What takes some setup:

• Batch processing workflows
• Custom label formats
• ILS import configuration

Here's my suggestion: Start small.

Pick 10 books that just arrived. Look up the MARC records yourself. Generate labels yourself. Process those 10 books from scratch.

You'll learn more in that one experiment than in hours of reading about it.

The Bigger Picture

This isn't just about saving money. It's about independence.

When your cataloging data comes from a vendor, you're dependent on that vendor. When your processing specs live in their system, you're locked in. When your entire acquisitions workflow runs through one company, you're one corporate decision away from chaos.

Four thousand libraries just learned that lesson the hard way.

The tools to do this yourself have existed for years. What's new is that AI removes the technical barrier. You don't need to understand APIs or write code. You describe what you need in plain English, and AI builds it.

That's the liberation. Not that the technology is new - it isn't. But now you can actually use it.

Want to go further? If you're ready to explore truly independent systems, check out Koha and Evergreen - open source ILS platforms used by thousands of libraries worldwide. Library Journal has good coverage of the open source library movement.