The specification has a price too: what a bulk Malbec programme actually requires

An illustrative request. Malbec at 12.5% ABV, with intense fruit, structure, integrated oak, a given price and enough volume to sustain a programme.

Those attributes can sit together perfectly well. What is still missing is what each one means, which are essential, and how they will be verified together.

When a buyer asks for Malbec at 12.5%, what else has to be settled for the sample to become a programme?

In the UK projects we work on, the alcohol level is set with its tax consequence in mind. Hitting the figure is one part of the job. The other challenge is preserving structure and power within that target, and that gets decided in the vineyard and the cellar.

Worth stating plainly: a wine at 12% is not the same as one at 13.5%. They are different propositions, not rungs on a single quality ladder. The aim is not to replicate a higher- alcohol wine with less alcohol, but to make a convincing wine at its own alcohol level.

Preserving structure within the target

The starting point comes before fermentation, and one figure does not settle it.

The Australian Wine Research Institute distinguishes ripeness of pulp, skin, phenolics, aromas and seeds, and notes that aromatic maturity does not necessarily coincide with sugar maturity. What counts as ripeness depends on the style being sought.

SourceAWRI, Defining berry ripeness, 2010

Which is why "pick earlier" is not a solution in itself. An earlier harvest lowers sugar, but it also changes the balance of other maturity components, and the result depends on which ripeness was being waited for. There is Argentine research on exactly this. A 2020 paper in OENO One tested nine combinations in triplicate on Malbec: two harvest dates — at 13.0% and 14.5% potential alcohol — partial replacement of the later harvest's juice with a low-alcohol, high- acidity Sauvignon Blanc wine, and different fermentation strategies.

The authors report up to 1.4 percentage points less alcohol and 0.3 units lower pH against the late-harvest control, with favourable chemical and sensory results under the conditions tested.

SourceFanzone and colleagues, OENO One 54(4), 1041-1058, November 2020

It matters to read this precisely. The intervention involved a second variety and is experimental work, not a Corbeau procedure nor a recipe transferable to any wine. What it establishes is that there is room for joint development, not that any combination of strength and profile is achievable.

There is also a reason why a single figure falls short. An AWRI project completed in 2017 examined how tannins, pH, alcohol and polysaccharides interact, with effects on texture, warmth and astringency. Alcohol does not act alone: it is part of a matrix.

SourceAWRI, project 3.1.4, completed June 2017, summarised by Wine Australia

A tannin or extract figure does not replace assessing the finished wine. The effect depends on composition, method and context.

From a number to a decision rule

"We're looking for 12.5%" leaves four different questions open, and they are worth closing before samples are selected.

• Is it the centre of a style range? Then it allows variation in both directions.
• Is it the maximum acceptable? Then it allows no variation upwards, and that changes which lots qualify.
• Is it the ABV sought for the finished product? A production target, which may not match the measured result.
• Is it the figure intended for the label? It may coincide with the target or the analytical result, but its declaration has to meet the applicable rules and does not replace the programme's acceptance criterion.

The four are not mutually exclusive: one project can have both a target and a maximum. What differs is the function of each figure and how it gets verified. Settling this prevents one party treating as compliant what the other considers outside the brief. Then comes how it gets measured. For its LM23 method, alcohol by distillation and densitometry, the laboratory associated with the AWRI declares an expanded uncertainty of ±0.1% v/v at 95%. That is the declared performance of that laboratory and that method, not a general industry tolerance.

SourceAffinity Labs, uncertainty of measurement for analytical laboratory methods, June 2025,

method LM23

A hypothetical example shows why this matters. Result 12.48% vol., declared expanded uncertainty of ±0.10 points, contractual maximum 12.50%. The interval runs from 12.38 to 12.58.

12.48 ± 0.10 → 12.38 to 12.58 · contractual maximum 12.50

Neither approval nor rejection follows automatically from that. The decision depends on the agreed rule and the applicable requirement. Uncertainty does not authorise exceeding a limit, is not a correction subtracted from the result, and does not guarantee where the true value sits.

International guidance on statements of conformity explains that the decision rule must correspond to the applicable requirements and be documented before the work. In a buyer's terms: the rule is identified and, where applicable, agreed before measuring, not after seeing the number.

SourceILAC-G8:09/2019, section 7

And four things are worth keeping apart: the contractual maximum, the label tolerance, the analytical uncertainty and the fiscal determination. Each answers to different criteria and requirements.

Alcoholic strength is defined at a reference temperature of 20 °C; the specification should also identify the analytical method used. The OIV provides for pycnometry and electronic densimetry on the distillate.

The specification is the whole set

The requirements need to be assessed together. Four dimensions are worth resolving before the first sample is requested.

An indicative framework for organising the brief; values and terms are agreed for each project.

There are also words that travel between the two sides without meaning the same thing. "Structure" can mean concentration, body, texture or length: worth asking which. "Power" does not automatically mean more alcohol or more extraction. "Intense fruit" requires a shared definition of the fruit character sought, and of the characters to be avoided. "Integrated oak" describes an outcome; if there are restrictions on treatment, they get stated separately. And "entry plus" or "high end" are commercial references, not analytical limits.

Negotiable preferences are worth separating from mandatory requirements. Safety, legality and authenticity are not variables that price can offset.

Availability is counted across the whole set

A hypothetical example, unrelated to any real stock. The brief calls for 120,000 litres of a defined profile, at a maximum of 12.5% vol., available within the project window. Four independent lots are assessed, with no subsequent treatment or blending.

Illustrative example. It does not represent Corbeau stock, nor does it demonstrate market scarcity.

There are 240,000 litres among the candidates, but only 60,000 meet every requirement at once. Availability is not added up before checking it against every requirement.

That result does not say the brief is technically impossible, or that other suppliers do not exist. It says that within this set, there is no basis for committing 120,000 litres.

The gap between the 120,000 litres requested and the 60,000 that comply forces a decision on how to complete the programme, and that decision can change both cost and schedule.

And it opens concrete decisions: find more candidates, cost out a development with its lead time, revisit a preference if it turns out to be negotiable, or move the calendar if the account allows it. If the alcohol maximum is mandatory, exceeding it to gain volume is not a solution. And averaging the strength of two samples does not produce a validated product.

When the specification changes the price

There is no universal premium for each requirement: its economic impact depends on the available wine and any development needed to meet the full brief. An additional constraint may not change the cost at all if the wine that meets it was already available; where it excludes candidates or calls for development, it can change selection, treatment, volume, lead time or cost. The consequence gets priced on the specific combination, not on the requirement in isolation. Three components are worth distinguishing in that conversation.

• Available wine that already complies. It may avoid additional development; the scope of the offer has to state what selection, validation and preparation it includes.
• Additional development required. Carries its own cost and lead time, best priced separately.
• Resources committed to repeating the profile. According to the terms actually agreed.

The question is not what it costs to drop half a degree, but which option meets the whole set of objectives and what commitments it requires.

The tax incentive behind the alcohol level

Commercial interest in particular strengths in the UK has a concrete explanation. Since February 2026 Alcohol Duty is £30.62 per litre of pure alcohol for wines between 8.5% and 22%, so half a point of difference comes to £0.1531 per litre of wine, before other effects.

SourceHMRC, Alcohol Duty rates, in force from 1 February 2026

That explains an incentive. It does not establish that the whole UK market buys at 12% or 12.5%, nor does it license choosing the declared figure for tax convenience. We set out the full calculation in an earlier Journal piece.

What to settle before the first sample

Five definitions organise everything above.

• What the ABV figure means. Target, maximum, production objective or label figure.
• Which sensory attributes are mandatory. And which are preferences open to discussion.
• By which method and rule acceptance is decided. Agreed and documented before measuring.
• Which reference identifies the profile. A sample or a shared description, not an adjective.
• What volume of that profile, and in what window. Because availability is counted across the set, not requirement by requirement.

Samples are best assessed blind, in comparable conditions, with observations recorded before they are discussed, and compared against the agreed profile rather than a personal preference. We develop this in the piece on continuity across vintages.

SourceAWRI, Practical sensory evaluation considerations

• In the UK projects we work on, the alcohol level is set with its tax consequence in mind. From the vineyard and the cellar we work to preserve structure and power within that target, through pruning, soil management, irrigation and picking date. We have achieved results we rate very highly, without confusing that work with a promise that a wine at 12% will be the same as one at 13.5%. The reference is the wine the project needs and the profile that gets agreed.