RESEARCH / CASE STUDY 01
From gift shelf to
global supply chain.
A conversation at Brookline Booksmith led me to a question about Barnes & Noble’s gift merchandise. I used public records to see what I could learn.
WHERE IT STARTED
A conversation at Brookline Booksmith sparked the research.
When I met with Brookline Booksmith in March, I learned how important gift baskets were to the store’s profits and how rising costs were making them harder to put together. That conversation made me curious about how a larger bookstore like Barnes & Noble sourced its gift merchandise.
I looked for what public import records could show. I gathered individual records, standardized inconsistent supplier names and product descriptions, and checked whether a few very large shipments were driving the results. The analysis offers a starting point for purchasing questions, not a direct comparison of the two businesses.
These figures describe a curated set of public ocean-import records. They are not Barnes & Noble’s full supply chain.
01 / BUILDING THE DATASET
Messy records into a usable sample
I assembled 30 records from public-facing trade-data pages, including ImportGenius, ImportInfo, ImportKey, i-trade, Seair, Trademo, and Zauba. I included records whose descriptions explicitly mentioned gift, stationery, craft, wrapping, drinkware, plush, puzzles, or related bookstore merchandise. Each row carries a source URL for inspection.
Supplier spellings were normalized for analysis while the original text remains in the CSV. I flagged descriptions containing multiple merchandise types, then calculated weight shares, supplier concentration, shipment-size comparisons, and sensitivity checks. Barnes & Noble’s public vendor materials supplied process context, not private transaction data.
02 / WHAT THE SAMPLE SHOWS
Concentration, with a crucial outlier check
The observed weight is concentrated by recorded source country and by supplier. Removing the largest shipment leaves the country pattern strong, while the leading supplier’s share falls substantially.
“China-linked” refers to the source-country field in these records, not confirmed manufacturing origin.
FORM + PAC represents 45.8% of observed weight; the top three groups total 63.9%.
The geographic concentration persists in this check. Supplier concentration is more sensitive to the outlier. Neither result estimates the full company-wide pattern.
03 / ANOTHER OPERATIONAL CLUE
Multiple product types travel together
Twenty-one of 30 records list multiple types of merchandise. These mixed-product records account for 67.8% of known sample weight. Their median known weight is 8.2 metric tons, compared with 3.1 metric tons for single-product records.
This is consistent with a system that can consolidate varied gift merchandise into larger inbound moves. The records alone cannot show freight savings, purchasing terms, or how items were distributed among stores.
The median comparison is descriptive, not a causal estimate.
04 / LOCAL APPLICATION
Questions an independent bookstore could ask
For a store such as Brookline Booksmith, the value is in adapting the method to its own private records. Without access to its internal data, I cannot make a numerical comparison with Barnes & Noble or claim how Booksmith currently buys.
How much purchasing depends on the top few vendors?
Which gift categories have the most volatile landed costs or lead times?
Could some small orders be combined without losing the local curation customers value?
A bookstore could answer these with supplier invoices, order dates, freight charges, sales, and margins held privately. The project includes a blank comparison template for that future analysis; it contains no Brookline Booksmith confidential information.
05 / METHODS & LIMITS
What these records cannot establish
The 30 verified records form a curated public sample, not a random or exhaustive census. Public aggregators differ in coverage and naming, and two rows lack usable weights. All weight percentages use the 28 weighted rows and refer to manifest shipment weight, which need not equal net product weight.
Recorded source country does not prove manufacturing origin. These records do not reveal unit costs, margins, tariffs paid, demand forecasts, store-level purchasing, or whether any item went into a specific gift basket.
The apparent July–September 2026 cluster should not be interpreted as seasonality: a partial public sample cannot separate real ordering patterns from changes in data visibility. The mixed-product comparison is descriptive and does not establish savings from consolidation.
I also have no access to Brookline Booksmith’s confidential sourcing or financial records. The questions above are potential applications, not findings about that business.
READ MORE OR INSPECT THE WORK
A short report, then the underlying data
I wrote a plain-language report for Brookline Booksmith that brings the findings and their limits together. The CSV includes row-level source URLs; the full project includes the analysis and methods.