Every January, Amazon's musical-instrument reviews outrun the December that preceded them — the sound of a million gifts being unboxed. And the size of each instrument's echo tells you exactly which ones are gifts and which ones are tools.
Retail data has an obvious December story: everything spikes. But a product review is not a purchase — it's a purchase plus an experience, written after the box is opened, the strings are tuned, and somebody has actually tried to play the thing. That delay turns Amazon's review timestamps into something more interesting than a sales calendar: a record of when products enter people's lives.
For musical instruments, the answer is January. In every year from 2015 to 2020, January beat the December that preceded it — by 9% to 32% — and together the two months run about a quarter above an ordinary two-month share of the year. The gifts bought in December echo through the review system for weeks: the keyboard unboxed on the 25th, attempted on the 26th, and reviewed on the 3rd.
Now split the same calculation by subcategory and the winter bulge becomes a diagnostic. Define a gift index: reviews in December + January relative to an average two months of that subcategory's year. An index of 1.0 means the winter holidays don't matter; higher means the product is, to some measurable degree, a present.
The ranking reads like a theory of gift-giving. At the top: keyboards (1.58×) and ukuleles (1.52×) — affordable, beginner-friendly, heavily marketed as first instruments. Guitars and drum kits follow. These are objects that carry a wish: the giver is buying somebody a new identity, or at least a new hobby.
At the bottom: studio recording equipment (1.13×) and live-sound gear (1.12×). Nobody gift-wraps a mixing console. These are tools, bought by people who already know exactly what they need, at the moment they need it — a demand curve with no December in it. Microphones (1.18×) sit tellingly close to the tool end: even in the giftiest weeks of the year, mic buyers are mostly buying for themselves.
It's a neat example of how much structure hides in a timestamp. Without a single survey question, the review calendar sorts an entire product category into presents and professional supplies — and quietly explains why instrument makers' fortunes swing with the holidays while the pro-audio business barely notices them.
Everything so far infers gifts from a calendar. But the reviews are text, and the text can testify directly. Two checks. First, count the share of reviews that actually say gift or Christmas, month by month:
Second, drop the word lists and let the corpus speak: score every word by how statistically distinctive it is of January reviews versus the rest of the year. No dictionary, no assumptions — just the words that January over-uses:
That third-person grammar is the strongest confirmation the gift-index story could ask for. The winter reviews aren't just more numerous — they are written by different people in a different role: givers reporting on recipients. The instrument bought in December is reviewed in January by someone watching somebody else play it.
Source: Amazon Reviews 2023 (McAuley Lab, UCSD), Musical Instruments category. Window: 2015–2019 pooled (pre-COVID; the 2020 shipping crunch and lockdown boom would contaminate a seasonality baseline). Figure 1: share of pooled reviews by calendar month. Figure 2: per subcategory, (Dec + Jan volume) ÷ (annual volume ÷ 6); subcategories with >8,000 reviews in the window. Reviews are dated by writing time, which lags delivery — the mechanism behind the January echo. Figure 3: share of review texts (title + body, lowercased) matching \b(gift|gifts|gifted)\b or \b(christmas|xmas|santa)\b. Figure 4: Monroe–Colaresi–Quinn log-odds with informative Dirichlet prior (Monroe et al. 2008), January vs. all other months; tokens split on non-word characters, length > 2, no stemming or stopword removal (the informative prior shrinks function words), vocabulary floored at 500 total uses; top 16 by z shown. Computed in DuckDB 1.5.4 (scripts/47_text_enrichment.py).
Reviews measure review-writing, not sales; if gift recipients review at different rates than self-buyers, the index mixes propensity with volume (the ranking is robust to this as long as the mix is stable within subcategory). Amazon's review-solicitation emails, which nudge reviews a fixed interval after delivery, reinforce rather than create the echo. In Figure 4, five and stars are an honest artifact: they come from Amazon's default “Five Stars” title on short reviews, which January's casual reviewers use disproportionately — itself evidence of who is reviewing that month.
Data © McAuley Lab, research release — cite Hou et al. (2024), arXiv:2403.03952. Article text and figures: CC BY 4.0.