01
Every export names fields differently.
A source may use product_title, unit_price, inventory_qty, or another internal vocabulary the destination does not know.
Map source headers to the exact BoundBot product fields.
CSV chatbot
Turn an existing export into a product source you can inspect.
Upload or paste CSV or TSV data, map unfamiliar headers, review the import, and use the resulting products in customer conversations.
CSV product import
Review before catalog write
northstar-products.csv
48 rows · 7 columns · 18 KB
Header mapping
product_titleName
unit_pricePrice
inventory_qtyStock
collectionCategory
Validation report
47 products ready
47
Ready
1
Review
0
Blocked
Row 18 · missing price
Review“Travel Mug · 16oz” can import after the required field is corrected.
Conversation preview
Do you have a travel mug under $30 that is currently in stock?
The 12oz Commuter Mug matches the imported catalog: $24, insulated, and currently in stock.
47 products ready · 1 row to reviewImport ready rows
Upload or paste
CSV, TSV, or text input
Map headers
Strict or AI-assisted mapping
Review issues
See invalid rows before import
Use the catalog
Power product conversations
The “we already have a CSV” problem
An export is portable. It is not automatically usable.
Real CSV files arrive with custom headers, partial rows, variant structures, and values that need to become reliable product fields before a bot should answer from them.
02
One malformed row can hide in a large file.
Missing names, inconsistent numbers, or variant mistakes are easy to miss until a customer asks about the product.
Review validation and rejected rows before trusting the import.
03
Import success is not answer quality.
A technically valid product record can still lack the use-case, compatibility, or description customers need to decide.
Test the customer question after the catalog exists.
A visible CSV pipeline
Load, map, validate, use.
BoundBot supports strict product headers and an optional mapping flow for unfamiliar exports, including structured variant imports.
Live decision path
Load
Upload the file or paste the rows.
Bring in CSV, TSV, or plain delimited text up to the supported file limit and preview the source data.
Control before catalog write
See how the file becomes product records.
A dependable import exposes the file, header interpretation, variant logic, and row outcomes instead of hiding everything behind a success toast.
Use a canonical BoundBot export when you want direct re-import with variant structure intact.
Use mapping when the source headers or row grouping come from another system.
Review the generated products and test customer language before launch.
CSV import stack
Sources the team can maintain
Source file
CSV, TSV, or pasted text
Header mapping
Source columns to product fields
Row validation
Required fields and transforms
Product catalog
Conversation-ready records
Flexible input, structured result
Use the export you have without treating it like a black box.
The CSV flow supports a fast strict import and a more deliberate mapping path when the data model comes from somewhere else.
01
CSV and TSV ingestion
Upload a file or paste delimited data directly into the product import experience.
02
Custom and variant mapping
Match arbitrary columns to product fields and describe row grouping when each row represents a variant.
03
Catalog-powered conversations
Use imported products across product finding, sales guidance, and support answers in connected channels.
Prepare the first export
Start with a small file you can verify by eye.
Prove the mapping and customer answer on a representative sample before importing the full catalog.
Export a representative sample.
Include normal products, one edge case, and variants if the real catalog uses them.
Choose strict or mapped import.
Use canonical headers for direct import or generate and review a mapping for custom columns.
Inspect product records.
Confirm names, descriptions, prices, stock, categories, images, and variant structure after parsing.
Ask catalog questions.
Test price, availability, fit, comparison, and compatibility questions before publishing the experience.
Which file types can I use?
The product import accepts CSV, TSV, and text-based delimited files. You can also paste the source data directly into the import form.
What if my column names do not match BoundBot fields?
Generate a mapping and review how each source column maps to product fields. A hint can describe special structure such as rows grouped into variants.
Can I re-import a BoundBot product export?
Yes. Canonical BoundBot product CSV exports can be imported directly, including supported variant structure.
Does importing a CSV create a chatbot by itself?
No. The import creates structured product data. Connect that catalog to a configured bot and customer channel to use it in conversations.
Continue exploring
Related BoundBot routes
Google Sheets Chatbot
Use Google Sheets as a lightweight source for chatbot knowledge, product data, and repeatable support answers.
Product Catalog Chatbot
Use a product-aware chatbot to answer catalog questions, availability questions, and buying questions from structured product data.
FAQ Chatbot
Build a FAQ chatbot that responds from structured knowledge instead of vague model memory.
Use the export you already have
Map the file. Review the rows. Put the catalog to work.
Start with a small CSV and verify the complete path from source column to customer answer.