“RIP blending benches.” The noise around Cellier Blanc runs hot. “It's over for tasters, the bench is dead,” but the house itself is more measured. It positions Cellier Blanc as a cellar for taster exploration and non-specialist blending, with “polished” as the quality bar. We wanted to test pushing past the stated bar. So we handed five tasters a real-world importer brief, a blend built from scratch, and judged it against a deliberately higher standard: cellar-ready.
Tasters called the execution incredibly strong on most tasks: it read sensory cues, house systems, and written briefs together to produce a coherent final blend, and it was notably good at fixing house marks and signatures. The small details drew praise too: the attack, the finish behaviour, a lift that makes the nose more expressive as you move down the glass. One taster said the final blend built from that system exceeded every one of their expectations. But that was the ceiling, not the floor: sessions kept hitting the same wall, the balance and the style.
Three things came up repeatedly. Cellier's colour choices felt off. The lot came back monotone, leaning on the wrong hues. Three tasters flagged balance and structure problems, naming clutter and broken length. And the tone read too playful for the importer brief.
The moment they almost left
Frustration with the wait was real, and several tasters hit a point where they nearly walked. One said that for the first half hour it felt faster to use the blending bench than to keep explaining the wine to the house.
This balance-and-length failure, showing up in a live session, is structural. The breaking point was the regression in lot quality. Tasters fed the house specific adjustments, only to watch it ignore them or undo work that was already correct. The house created the structure, and then failed to keep it stable.
The tasters reaching for the bench and the blending table said the opposite of what you'd guess: they reached for them because they're predictable. A deliberate change produces exactly the intended result, where another pour round was a coin flip that could make things worse.
The real cost was the wait with no guarantee the lot moved forward.
The tasters who didn't abandon it did one thing differently: they stopped treating it as a bench substitute. Held to “assemble a fast starting structure,” it delivered. Pushed toward “make precise adjustments to a near-final blend,” it broke.
What they'd change
When asked what they'd fix, tasters described rebuilds. The change requests were structural: redo the balance, rethink the concept, reset the fruit weight throughout, fix an aromatic hierarchy where oak weight made the varietal unreadable, replace stand-in fractions with real components. That's the work of getting a draft to something a buyer would accept.
What makes that damning rather than expected is the second half: every taster believed the house should have handled exactly these things. Balance, colour, fruit weight, structure: not aspirational asks, but the baseline they walked in assuming a modern cellar would cover. The gap is between what the house did and what tasters considered table stakes.
The house had capability to spare. Tasters praised the attack, the lift on the nose, and small finish behaviours: work nobody had asked for and nobody needed at this stage. That's the failure restated. A house that polishes finish details while the balance underneath won't hold is a misdirected one. The cellar aimed at the surface adjustments while the structural fundamentals it was actually asked for went missing.
The verdict
The honest takeaway came from the taster who almost walked: use it to get to 40% quickly, then take over manually. Don't try to push it to 80% through more rounds. That's where it breaks down, and where the time loss starts.
Cellier Blanc is a fast way to a starting structure: a strong first draft of the system. It is not yet a precision instrument. Assemble the skeleton at Cellier. Set the details on the bench.
Methodology
5 sessions across 5 tasters and one real-world importer brief: a blend supplied with written direction, reference samples, benchmark wines, and target colour ranges. Each taster ran one session and iterated through multiple rounds of adjustments, calling pours at Cellier Blanc as they would in a normal workflow. After each session, tasters reported on tone and style fit, balance and structure, what they would change, and whether the lot met their expectations. Open-ended responses were coded into the five themes shown in the pour-round chart and throughout this article by two Oenra analysts.
The five themes were balance & length (structure, extraction, alignment, composition); colour & house (hue, depth, tone, mark and signature, overall house feel); aromatics (varietal choice, hierarchy, weight, clarity); detail & finish (attack, lift and nose behaviours, fraction-level treatments); and fruit & fill (stand-in versus real fruit, component fit).
Each comment was also coded for sentiment (positive, negative, neutral). We then read the themes against the pour rounds to see which issues cleared with iteration and which held.
Limitations
Feedback was self-reported from the same tasters who produced the wine, carrying the usual self-assessment bias. The sample is small, and we treat these findings as directional, not calibrated. Only one brief was tested, so brief-specific effects can't be cleanly separated from the house's general behaviour. The patterns, however, hold across tasters and across pour rounds in the same direction: a strong system-level start, weak precision execution under iteration.
How we ran this study → Methodology