A chocolate bar is a spreadsheet wearing a bright wrapper. Cocoa, sugar, milk powder, vegetable oil, foil, paper, freight, energy and labor all arrive with their own weather, politics and price. When a supplier says the bill must rise by 11 percent, a buyer has two options: nod, or open a better spreadsheet. Expana sells the better spreadsheet - plus the data, forecasts and independent benchmarks needed to defend it.
The company occupies a peculiar but consequential corner of business information. It is a price reporting agency, or PRA, focused on agrifood. Its reporters gather evidence from physical markets where trades are private, products are inconsistent and a tidy exchange price often does not exist. Expana turns that evidence into reference prices. Around those numbers it has built software for forecasts, cost modeling, market news, supply and demand, crop conditions and alerts.
That makes the product useful to the people who buy cocoa rather than chocolate, chicken rather than sandwiches, resin rather than bottles. Procurement and category teams use it to challenge quotes. Manufacturers use it to understand margins. Traders and finance teams watch risk. Consultants and analysts pull the data into their own models. More than 3,000 organizations use the service, according to Expana, including retailers, producers, foodservice groups, manufacturers and financial firms.
01 / A company younger than its archive
The 168-year-old startup problem
Expana arrived as a name in 2024. Its ancestry arrived much earlier. Urner Barry traces its reporting history to 1858, when collecting market intelligence meant talking directly to New York merchants. The Urner Publishing Company began a daily price circular in 1882. In 1980, it launched COMTELL, moving real-time agricultural market data onto screens. Mintec, founded in 1982, built its own position in raw-material prices and procurement intelligence.
The modern company was assembled through acquisition. Mintec bought forecasting specialist Kairos Commodities in 2021, French data provider CommoPrices in 2022 and AgriBriefing in 2023. That last transaction added Urner Barry, Feedinfo, Stratégie Grains and Tropical Research Services. Five Arrows Principal Investments backed Mintec, with Synova Capital holding a significant minority investment. The prices paid were not announced; the commercial bet was plain enough. Specialist data sets become more useful, and more defensible as subscriptions, when they meet inside one workflow.
The rebrand to Expana was an admission that a shelf of respected names was not yet a coherent product. Customers dealing with a finished food do not care which corporate family owns the egg series and which one forecasts wheat. They want the components to speak to one another. The unified platform went live in 2025, pulling the acquired archives, analysis and modeling tools under one login.
“One absolutely critical version of the truth.”Sam Pearl, former senior buying manager at Tesco, on Expana's data
02 / Where the number comes from
A benchmark must survive an argument
Data volume is the easy boast. Expana advertises more than 41,000 time series across over 900 commodities and more than 70 million data points through its API. The harder claim is that a buyer and seller can use one of its proprietary benchmarks without either side feeling ambushed. That requires rules: precise product grades, regions, cargo sizes, delivery windows, currencies, contributor checks, peer review, complaint handling and corrections.
Expana says its methodology is aligned with the principles for price reporting agencies set by the International Organization of Securities Commissions. An external assurance review tests the policies and controls; the company completed its sixth consecutive annual review in January 2026. This is different from IOSCO certifying every number. It is assurance over the process that makes the numbers, an unglamorous distinction that matters when a price becomes a clause in a contract.
From a messy market to a usable decision
This reporting machinery is Expana's less fashionable advantage. An AI model can forecast a bad series with great confidence. It cannot cheaply recreate a contributor network, decades of market definitions and the institutional permission to publish a number that both sides will quote. The new software is valuable because the old reporting work sits underneath it.
03 / The product is a decision
From cocoa tick to boardroom answer
The platform wraps those benchmarks with forecasts, news, fundamentals, weather and crop analytics. Its cost models connect raw materials with packaging, freight, energy, labor and other expenses. A user can build a model of a finished product, watch which input moved and project the effect on cost for up to 18 months. Supplier quotes and third-party data can be added; results can travel by API into ERP, finance, procurement, Excel or business-intelligence systems.
In August 2026, Expana launched a batch of more than 500 Expana IQ forecasts across 16 top-level categories, increasing forecast coverage by 24 percent. The models are not presented as one magic algorithm. Expana's data scientists select statistical or machine-learning methods suited to each market, applying AI selectively and under supervision. That restraint is sensible: poultry, chemicals and herbs do not move for identical reasons merely because all can be put in a table.
An 18-month line of sight
Cost models connect changing inputs to a finished SKU, giving procurement and finance teams time to budget, renegotiate, hedge or reformulate.
The output is not meant to admire. It is supposed to change a meeting. A buyer can compare a supplier increase with the ingredients that should drive it. Finance can replace a flat inflation assumption with component forecasts. A product team can test whether changing a recipe or package shifts the economics. The Data Direct API matters here: market intelligence becomes far stickier when it appears inside the spreadsheet and dashboard people already open.
04 / When data becomes infrastructure
The price starts doing things
A benchmark becomes more interesting when money can settle against it. Many food commodities have no liquid listed futures contract, leaving companies to absorb volatility or hedge with an imperfect substitute. In 2025, Expana and StoneX announced OTC dairy derivatives referencing Expana's European dairy benchmarks. The first fat-filled milk powder contract became available in April 2026, with more products planned.
Three months later, French investment firm Quideos licensed the full range of Expana Benchmark Prices for tailored OTC risk-management services. Its pitch is almost comically practical: hedge the index and volume a food business actually uses, including products ignored by listed exchanges. The benchmark is no longer a fact on a screen. It is plumbing for a financial contract.
Expana is also willing to rent missing pieces. A 2026 partnership puts Fastmarkets' forest-products benchmarks into the platform, giving customers upstream prices for wood, pallets and recovered paper beside Expana's downstream packaging intelligence. Fastmarkets is an obvious competitor in commodity data, alongside S&P Global's Platts lineage, Argus Media and ICIS. Partnership makes sense when the customer values a complete workflow more than corporate purity.
The bit worth copying
Start with a costly recurring argument, not a dashboard. Define the disputed object precisely. Build a process that both sides can inspect. Then place the answer inside the tools where a decision already happens. Expana's sequence - trusted benchmark, connected model, embedded workflow - is useful well beyond commodities.
The model has limits. It works best where enough physical-market evidence exists, product specifications can be normalized and customers repeatedly make decisions large enough to justify an enterprise subscription. A thin market, a bespoke input or a small buyer with little negotiating power will get less value. Forecasts can clarify exposure; they cannot abolish weather, war, disease, tariffs or the human appetite for changing the recipe.
Nor does the long heritage remove integration risk. Expana must keep specialist expertise intact while making acquired datasets behave consistently, persuade users to trust a consolidated brand, and avoid letting automation outrun the evidence underneath it. The first thing that failed was not a price model but the old portfolio logic: separate brands and interfaces could not deliver the single view customers needed. The answer was the 2024 identity and the 2025 platform.
Under CEO Julie Harris, appointed in January 2026 after leading beverage-data company IWSR, Expana is pushing that platform further into forecasts and financial use. The result is neither a terminal for every commodity nor a generic AI oracle. It is narrower and, for its customers, more concrete: a way to ask what a thing should cost, why it moved, what happens next and whether anyone will agree on the answer.