The buyer who treated fresh produce as a forecast
Fresh produce is the hardest thing in retail to run well, because the product is destroying itself the entire time the business is making decisions about it. A delayed container is a total loss, and that single physical fact has produced a small number of operators who stopped treating the supply chain as a sequence of negotiations and started treating it as a forecasting problem. One of them was recognised in the Best in Business Awards 2025, in front of jurors who read the file for whether the analysis predated the result or had been assembled afterward.
Olga Maximchuk took the 2025 recognition as Woman Leader of the Year. She has more than twenty years in executive retail, with category-management work at X5 Retail Group between 2016 and 2021, which the submission credits with cumulative profit growth above 2.5 billion roubles and anti-crisis initiatives saving 150 million roubles a year, followed by logistics-cost work at O’Key that cut logistics cost by 7 percent, and the building of Vegberry, a controlled-environment agriculture operation in the UAE, grown from 2.2 hectares to more than 60, into a top-three vegetable supplier in that market with operations extending into Oman. Her work has been covered in Retail.ru and RBK. The figures here are as reported in the submission to BIBA 2025; the program published its criteria, but the individual scores remained private.
The intellectual core of the file is the X5 category work, because it is where the method was proven before it was exported. Machine-learning demand forecasting that incorporated weather, quarantine, and transportation-flow variables cut write-offs by 15 percent; an assortment-optimization model found that 6 of 27 tomato SKUs generated 80 percent of the category’s profit, and acting on that lifted inventory turnover by 35 percent. Those are analytical outcomes, and Maximchuk’s own framing of the work puts the emphasis there.
“The future belongs not to those who bargain better, but to those who analyze better,” Maximchuk said.
The Vegberry build is the same method applied where the stakes are highest, because the product is perishable controlled-environment produce supplying a market that does not tolerate gaps. At that scale, the demand model is upstream of the entire operation: it sets grower contract volumes, which set planting schedules and capacity, so the forecast is in place before the season and the season is run against it.
“These are not temporary difficulties — this is a new reality that requires rebuilding the entire business model. A single delayed container is not a financial loss; it is a direct loss of one hundred percent of the cargo value, because our product lives for only a few days,” Maximchuk said.
The model she built around that reality, which the submission calls UltraFresh 3.0, replaces the adversarial buyer-supplier relationship with co-investment and shared risk. The clearest instance is a Serbian berry deal: cold-storage capital expenditure funded in exchange for 30 percent of an exclusive premium variety’s harvest at a fixed price for three years. That is written into a contract, with the obligations on both sides fixed for the term, and it works only if both sides can see the same forecast.
“Our task is not simply to dictate terms, but to be a partner that helps the other side modernize. We run training seminars, share practices, and help with access to financing. We are interested in a strong, technology-capable partner — not a dependent supplier,” Maximchuk said.
On top of the operation sits a harder problem, grounding a claim about perishable production in data that a third party can verify. Vegberry’s blockchain traceability records the product from field to shelf, and the Vegberry Coin tokenizes agricultural assets, which only carries meaning if the audited operating record underlying it is real. The submission frames the whole arrangement as a chain in which no participant is independent of the others, with the Farmkey and Vegberry Solutions businesses built on the same ecosystem.
“The era of one side dictating is over. We are links in the same chain, and if one link collapses, everyone suffers,” Maximchuk said.
The juror best placed to read the measurement discipline behind a forecasting build was one who teaches exactly that. Iryna Smuk founded the agency BRKTHROUGH and the company Solmar, has been featured in Harper’s Bazaar, MC.today and Cosmopolitan, is a TOP Marketing Club Ukraine member, and has trained more than 130 professionals in her own SMIV Framework, a structured measurement methodology. Her test is whether the tracking predated the story.
“The submissions that stood out were the ones where someone had clearly been tracking their numbers from the beginning,” Smuk said. “Not because they were building a case for an award, but because that’s how they work. That habit shows up in the quality of the evidence.”
Smuk’s test is decisive on a perishable-supply file because there is nowhere to hide a reconstruction. A demand model that drives planting schedules leaves a continuous record, since the operation depends on it daily; a write-off reduction is visible only if the write-offs were being measured before the reduction. The 15 percent write-off cut and the tomato-SKU concentration are the kind of figures her framework expects to find instrumented in the operation and computed for no deck, because an operator who found them after the fact could not have acted on them in time to change the season. The 7 percent logistics-cost reduction at O’КЕЙ and the extension of Vegberry operations into Oman are the same instinct at different scales, each a number that only exists because the measurement was running before the decision that moved it. More than 130 practitioners have been trained through her agency work to look for exactly that seam, and a fresh-produce operation that scaled from 2.2 to 60 hectares without one is a file where the measurement was load-bearing the whole way. The BRKTHROUGH and Solmar businesses she built gave her the operator’s version of the same eye: a measurement she did not keep from the beginning is one she has watched fail to survive the questions.
The forecasting work is the part of the file that a confident summary never has to show. A demand model that drives planting schedules leaves a continuous record because the operation depends on it daily. The Serbian deal is the same logic in a contract: cold-storage capital placed before a single harvest arrived, against three years of fixed-price exclusive premium harvests, an exposure an adversarial buyer would never accept, and a forecast both sides have to be able to read for the deal to hold. Blockchain traceability and the Vegberry Coin tokenization sit on top of an audited operating record, which is the part that makes claims about perishable production checkable by someone outside it.
The cost of running a supply chain this way is the part the negotiating story leaves out. Building forecasting infrastructure, funding a supplier’s cold storage before a single harvest arrives, and tokenizing an asset whose value depends on an audited record are all expensive commitments made before the return is known. The compensation is that the resulting claims are unusually hard to dispute, because each one rests on a record that a counterparty keeps. The Serbian grower holds the three-year fixed-price contract. The retailers and food-service operators who place Vegberry among their top three suppliers are the instrument for that ranking, and it cannot be asserted without them. The blockchain ledger records the product as it changes hands that are not the operator’s. A file built like that survives a skeptical read because the operator chose, at cost, to put the proof outside her own control.
A top-three vegetable supplier position in a market is a buyer-side fact, set by the retailers and food-service operators who place a supplier there against their own procurement standards, and it cannot be asserted on a slide without them. The same is true of the write-off cut: a 15 percent reduction is a number that an operation either kept while it ran or reconstructed afterward, and the difference is visible on the first page. The operators in this cohort built their figures into the records their counterparties keep, which is why a thirty-five percent turnover lift tied to a six-of-twenty-seven SKU finding carries information that a claim of efficiency does not.
Other operators recognized in adjacent supply-chain and leadership categories worked the same territory the panel reviewed, among them Kuanysh Mustafin in business scaling. The work that lasts in perishable supply chains is the forecast that sets the contract before the season, kept because the operation could not run without it, and the partnership structure is written into a deal where a slide would have sufficed elsewhere. A year from now, the citations will have faded, and the procurement records will not have. Maximchuk’s supplier ranks will still sit in the buyers’ systems, the Serbian contract will still carry its fixed price, and the operators worth remembering from BIBA 2025 in this territory are the ones whose hardest numbers were being kept for the operation’s own sake long before there was a category to enter them in.