How to read these
Each recipe lists the prompt to paste, which tools and fields it exercises, the output you should expect, and the one caveat worth stating alongside the result. Prompts start with "Using ALT/FNDATA" so Claude reaches for the connector. Swap the brand, category, or date range freely; the same shape holds because every category shares one schema.
One habit to keep Ask Claude to filter to sold lots before quoting a price, and to flag the most recent one or two quarters as provisional. Both are built into the recipes below.
Family office
Family offices size and monitor a collection at the market level, the way they would any illiquid sleeve of the portfolio.
Collection allocation and liquidity
Family officeUnderstand what a family's collecting categories are worth at the market level, and how liquid each is, without pretending recent quarters are complete.
Using ALT/FNDATA, for the watches, jewelry, and handbags tables, give me the median sold price and the number of sold lots by year for the last five years. Filter to sold lots only, and note if the most recent year looks incomplete.
Now show the sell-through rate by year for watches: sold lots as a share of all offered lots. Keep unsold lots in the denominator.
Exerciseslist_tables, get_table_schema, then query_table per table with a status = sold filter and a sale_date range, sorted by date
Output shapeA per-category, per-year table of median price and sold-lot count, with a note on the latest period
CarryNewest quarters undercount; the most recent year is provisional
Concentration by brand and category
Family officeSee where a collection's market lives, so a family knows its single-name and single-category exposure before it grows.
Using ALT/FNDATA, in the watches table, show the ten brands with the most sold lots over the last three years, with each brand's sold-lot count and median usd_price_decimal. Sold lots only.
Now run the same for the jewelry and handbags tables so I can compare where value and liquidity sit across all three categories.
Exercisesquery_table on watches, jewelry, and handbags filtered to status = sold over a sale_date range, grouped by designer in the answer
Output shapeA ranked brand table per category, showing where demand and value concentrate
CarrySecondary-market activity, not the family's actual holdings; recent quarters provisional
How reliable are pre-sale estimates
Family officeJudge how close hammer prices land to estimates in a category, a proxy for how predictable that market is when you buy or sell.
Using ALT/FNDATA, in the watches table, for sold lots over the last two years where sale_estimates_high_usd_price is greater than zero, show the median ratio of usd_price_decimal to sale_estimates_high_usd_price by year. Sold lots only.
Repeat for jewelry and fine_art, and tell me which category tracks its estimates most closely.
Exercisesquery_table with status = sold and sale_estimates_high_usd_price greater than zero; the ratio is computed by Claude over the returned rows
Output shapeA by-year price-to-estimate ratio per category, a read on pricing predictability
CarryEstimates are forecasts, not paid prices; ratios use sold lots only
Hedge-fund analyst
Analysts treat auction demand as an alternative dataset: a high-frequency read on brand desirability that runs ahead of the primary-market numbers a company reports.
Luxury demand read-through to equities
Hedge-fund analystUse auction demand as an alternative signal on a listed brand's health, and tie it back to the parent company's ticker.
Using ALT/FNDATA, in the watches table, show quarterly sold-lot counts and median sold price for Rolex and for Patek Philippe over the last three years. Then tell me each brand's parent-company stock ticker from the data.
Rank the top ten brands in the watches table by median sold price over estimate, using usd_price_decimal divided by sale_estimates_high_usd_price, only where the estimate is greater than zero and there are at least twenty sold lots.
Exercisesquery_table on watches filtered by designer and status = sold across a sale_date range; stock_ticker for the parent link
Output shapeTwo quarterly series (count and median price) per brand, plus the parent ticker for each
CarrySecondary-market demand, not primary sales or revenue; recent quarters provisional
Spot a turn in brand demand
Hedge-fund analystDetect when a brand's secondary-market demand is accelerating or rolling over, quarter by quarter, before it shows up in reported results.
Using ALT/FNDATA, in the watches table, show quarterly sold-lot count and median usd_price_decimal for Audemars Piguet over the last four years. Sold lots only, and flag the most recent one or two quarters as provisional.
Overlay the median ratio of usd_price_decimal to sale_estimates_high_usd_price by quarter, where the estimate is greater than zero, so I can see demand and pricing power together.
Exercisesquery_table on watches filtered by designer and status = sold across a sale_date range, summarized by quarter
Output shapeA quarterly demand-and-pricing series for one brand, with the recency caveat stated
CarrySecondary-market demand, not revenue; the last one to two quarters are provisional
Relative strength across a peer set
Hedge-fund analystRank a peer group by pricing power to build a relative-value or pairs view on the listed parents.
Using ALT/FNDATA, in the watches table, for Rolex, Patek Philippe, Audemars Piguet, and Omega over the last three years, show sold-lot count and the median ratio of usd_price_decimal to sale_estimates_high_usd_price, where the estimate is greater than zero and there are at least twenty sold lots per brand. Include each brand's stock_ticker.
Rank the four by that ratio and tell me which parent tickers they map to.
Exercisesquery_table on watches by designer and status = sold; stock_ticker for parents; ratio and rank computed by Claude over the rows
Output shapeA peer-ranked pricing-power table with parent tickers attached
CarryA demand signal, not a valuation call; recent quarters provisional
Wealth advisor
Advisors put a defensible, sourced number next to a client's passion asset, whether the client is buying, selling, insuring, or simply asking what it is worth.
Client holdings valuation context
Wealth advisorGive a client a defensible market range for an asset they hold or want, grounded in comparable sold lots rather than a single data point.
Using ALT/FNDATA, find sold Hermes Birkin and Kelly handbags from the last three years. Show the distribution of usd_price_decimal, and the low, median, and high, filtered to sold lots. Note the auction houses represented.
Narrow it to lots whose item_title mentions crocodile or Himalaya, and show how the median shifts.
Exercisesquery_table on handbags with designer and item_title like filters, status = sold, grouped in the answer by vendor
Output shapeA price range (low, median, high) over comparable sold lots, with the houses that transacted them
CarryA market range from comparables, not an appraisal; excludes retail and dealer asking prices
Is now a good time to sell
Wealth advisorShow a client whether the market for their asset is rising or softening before they consign.
Using ALT/FNDATA, in the watches table, for sold Patek Philippe lots whose item_title mentions Nautilus over the last four years, show median usd_price_decimal and sold-lot count by year. Sold lots only.
Flag whether the most recent year is provisional, and note the range of auction houses that sold them.
Exercisesquery_table on watches with designer and an item_title like on the model, status = sold, summarized by year
Output shapeA multi-year price and volume trend for a specific model, with the recency caveat
CarryTrend from secondary-market sold lots; the latest year is provisional
A replacement-value range for a schedule
Wealth advisorGive a client a market-based figure to schedule an item for insurance, grounded in recent comparable sales.
Using ALT/FNDATA, in the jewelry table, find sold lots whose item_title mentions sapphire ring from the last three years, and show the low, median, and high usd_price_decimal. Sold lots only, and list the vendors.
Narrow to lots that also mention Kashmir or unheated and show how the range moves.
Exercisesquery_table on jewelry with item_title like filters and status = sold; low, median, and high over the returned rows
Output shapeA comparable-sales range to support a scheduled value
CarryA market range from comparables, not a certified appraisal; condition and certification vary by lot
Luxury-brand strategist
Brand and insights teams read the secondary market as the scoreboard for brand equity: how well products hold value, and where demand is rotating.
Competitive resale benchmarking
Luxury-brand strategistBenchmark a house's resale strength against rivals: how prices hold versus estimate, and where demand is trending across categories.
Using ALT/FNDATA, compare Cartier and Van Cleef and Arpels in the jewelry table over the last four years. For each, show sold-lot count by year and the median ratio of usd_price_decimal to sale_estimates_high_usd_price, only where the estimate is greater than zero.
Which vendors account for most of each brand's sold lots? Group by vendor and show the top five.
Exercisesquery_table on jewelry filtered by designer and status = sold; the ratio is computed by Claude over the returned rows
Output shapeA by-year comparison of demand (counts) and pricing power (price-to-estimate ratio) for two brands
CarryAuction resale behavior, not primary retail performance; recent periods provisional
Which model lines hold value
Luxury-brand strategistSee which of a house's model lines command the strongest resale, to inform product and pricing strategy.
Using ALT/FNDATA, in the watches table, for Omega over the last three years, group sold lots by the model line named in item_title (Speedmaster, Seamaster, Constellation) and show sold-lot count and median usd_price_decimal for each. Sold lots only.
Add the median ratio of usd_price_decimal to sale_estimates_high_usd_price per line, where the estimate is greater than zero.
Exercisesquery_table on watches by designer with an item_title like per line and status = sold
Output shapeA per-model-line demand and pricing comparison within one brand
CarryAuction resale, not primary sell-through; recent periods provisional
Where does resale demand sit
Luxury-brand strategistUnderstand which houses and channels move a brand's pieces, to target partnerships and authentication programs.
Using ALT/FNDATA, in the handbags table, for Chanel over the last three years, show the top ten vendors by sold-lot count, with each vendor's median usd_price_decimal. Sold lots only.
Which vendors show the highest median price for the brand, and how concentrated is volume in the top three?
Exercisesquery_table on handbags by designer and status = sold, grouped by vendor in the answer
Output shapeA vendor league table for one brand: where volume and the highest prices sit
CarryAuction and marketplace resale only; recent quarters provisional
Appraiser
Appraisers pull a clean, dated, sourced comp set into a valuation file in minutes instead of hours.
Comparable-sales evidence
AppraiserAssemble a clean, sourced set of comparable sold lots for a valuation file, with the auction house and date on every comparable.
Using ALT/FNDATA, in the watches table, find sold examples of Rolex Daytona reference 116500LN from the last two years. Return item_title, vendor, sale_date, and usd_price_decimal, sorted by sale_date descending, sold lots only.
Page through the next twenty results with offset so I have the full set of comparables.
Exercisesget_table_schema to confirm fields, then query_table on watches with designer, an item_title like on the reference, and status = sold
Output shapeA dated, sourced list of comparable sold lots ready to cite in a valuation
CarryComparables from secondary-market sales; verify condition and specifics against each lot
Comps filtered by condition language
AppraiserTighten a comp set to lots that share the subject's configuration, using the words that appear in the catalog title.
Using ALT/FNDATA, in the handbags table, find sold Hermes Kelly 28 lots from the last three years whose item_title mentions togo or epsom. Return item_title, vendor, sale_date, and usd_price_decimal, sorted by sale_date descending. Sold lots only.
Now split the set by whether the item_title mentions gold hardware or palladium, and show the median of each.
Exercisesquery_table on handbags with designer and multiple item_title like filters, status = sold
Output shapeA configuration-matched, dated comp list, optionally split by hardware or leather
CarryTitles describe but do not guarantee condition; verify each lot
Sanity-check a number against the market
AppraiserTest a proposed value against how comparable lots actually landed versus their estimates.
Using ALT/FNDATA, in the watches table, for sold Cartier Santos lots over the last two years where sale_estimates_high_usd_price is greater than zero, return item_title, vendor, sale_date, usd_price_decimal, and sale_estimates_high_usd_price. Sold lots only.
Compute the ratio of usd_price_decimal to sale_estimates_high_usd_price for each and give me the median.
Exercisesquery_table on watches by designer and status = sold, returning sale_estimates_high_usd_price; the ratio is computed over the rows
Output shapeA comp table with realized-to-estimate ratios to frame a valuation
CarryEstimates are forecasts; sold lots only, and condition varies
Financial journalist
Reporters stand a market claim up or knock it down with sold-lot evidence, and avoid mistaking ingestion lag for a trend.
Market-trend fact-checking
Financial journalistStand up or knock down a market-trend claim with sold-lot evidence, and avoid mistaking ingestion lag for a real move.
Using ALT/FNDATA, is the classic-car market cooling? In the automobile table, show sold-lot counts and median sold price by quarter for the last three years. Flag whether the most recent quarters may be incomplete due to ingestion lag.
Break the most recent year down by vendor so I can see which houses drive the totals.
Exercisesquery_table on automobile filtered to status = sold across a sale_date range, summarized by quarter
Output shapeA quarterly trend of volume and median price, with the recent-quarter caveat stated
CarryA dip in the last one to two quarters is likely ingestion lag, not a market decline
Top lots of the year
Financial journalistFind the headline sales in a category for a year-in-review or a record-price story, with houses and dates.
Using ALT/FNDATA, in the jewelry table, show the twenty highest usd_price_decimal sold lots of the last twelve months. Return item_title, designer, vendor, sale_date, and usd_price_decimal, sorted by price descending. Sold lots only.
How many of the top twenty were sold by each house?
Exercisesquery_table on jewelry with status = sold, sorted by usd_price_decimal descending, limit 20
Output shapeA ranked top-lots list ready to cite, plus a house tally
CarrySecondary-market results; very recent sales may still be landing
Which house leads a category
Financial journalistCompare auction houses' share of a category to support a competitive-landscape piece.
Using ALT/FNDATA, in the automobile table, over the last two years show the top ten vendors by sold-lot count, with each vendor's median usd_price_decimal. Sold lots only.
Which vendor has the highest median price, and which has the most volume?
Exercisesquery_table on automobile with status = sold, grouped by vendor in the answer
Output shapeA house league table by volume and median price for the category
CarryCoverage is secondary-market; recent quarters provisional
Dealer and trader
Dealers and secondary-market traders price inventory, spot where a segment is mispriced, and time sourcing and consignment.
What should this trade at
Dealer and traderGet a fast, defensible market range for a specific reference or model before making an offer or setting a list price.
Using ALT/FNDATA, in the watches table, find sold Rolex Submariner reference 126610LN from the last eighteen months. Return item_title, vendor, sale_date, and usd_price_decimal, sorted by sale_date descending. Sold lots only, and give me the low, median, and high.
Page through the next twenty with offset so the range reflects the full set.
Exercisesquery_table on watches with designer and an item_title like on the reference, status = sold; low, median, and high over the rows, with offset for depth
Output shapeA recent, dated range for one reference to anchor a buy or list price
CarrySecondary-market sold lots; condition and completeness such as box and papers move price
Where is the segment heading
Dealer and traderSee whether a model's market is firming or softening quarter by quarter, to decide whether to hold or move inventory.
Using ALT/FNDATA, in the handbags table, for sold Louis Vuitton lots whose item_title mentions Capucines over the last three years, show median usd_price_decimal and sold-lot count by quarter. Sold lots only, with the latest quarters flagged provisional.
Break the most recent year down by vendor.
Exercisesquery_table on handbags with designer and an item_title like, status = sold, summarized by quarter
Output shapeA quarterly trend for a model to guide hold-or-sell timing
CarryThe recent one to two quarters undercount; not a guaranteed forward price
Insurance and risk
Underwriters and claims teams set scheduled values and test claimed values against the market for collectible risks.
Set a scheduled value
Insurance and riskAnchor a scheduled value for a collectible on recent comparable sales rather than a single receipt.
Using ALT/FNDATA, in the jewelry table, find sold lots whose item_title mentions diamond eternity band from the last three years, and show the low, median, and high usd_price_decimal with the vendors. Sold lots only.
Narrow to lots that also mention a carat weight at or above three carats and show how the range moves.
Exercisesquery_table on jewelry with item_title like filters and status = sold; low, median, and high over the rows
Output shapeA comparable-sales range to support a scheduled value
CarryA market range, not a certified appraisal; certification and condition vary by lot
Is a claimed value in range
Insurance and riskCheck whether a claimed or declared value sits within the recent market for comparable lots.
Using ALT/FNDATA, in the watches table, find sold Patek Philippe lots whose item_title mentions Aquanaut from the last two years. Return item_title, vendor, sale_date, and usd_price_decimal, sorted by sale_date descending. Sold lots only, and give the median and the tenth to ninetieth percentile range.
Flag any lots more than fifty percent above or below the median so I can review outliers.
Exercisesquery_table on watches by designer and an item_title like, status = sold; the distribution is computed over the returned rows
Output shapeA market distribution to place a claimed value in context
CarrySecondary-market comparables; condition, papers, and provenance affect value