ParallaxAdvisory
Back to selected work
Case 06 — What's Your Whisky

Integrating AI across the full marketing mix.

Premium whisky is one of the highest-value categories in drinks — yet its volume and value were capped by a set of stubborn barriers. AI was the way to reduce them.

Two barriers were holding the category back. For consumers, premium whisky is among the most expensive sub-categories in spirits — bottles frequently cost hundreds, and rare bottlings many thousands, of pounds — so a buyer has to feel confident they'll actually enjoy it before they commit.2 For the business, Diageo had only limited consumer flavour-preference data across its markets, so decisions on which of its many whiskies to sell, to whom and where ran largely on instinct. The question that framed the work: how could one platform give consumers the confidence to buy now — driving revenue in the near term — while at the same time generating consumer preference data at enormous scale, the engine of long-term value?

RoleRegional Senior Brand Manager
ClientDiageo · Reserve
ScopeGlobal
Year2019–20
ProgrammeFIELD · DEPTH
ProofPublic · 1M+ consumers worldwide
01The problem

Two barriers, one root: limited flavour data and low buyer confidence.

Two barriers sat on either side of the same category. The business had only limited consumer flavour-preference data across its markets, so decisions on which whiskies to range, for whom and where ran largely on instinct. And the consumer — facing a premium whisky that can cost hundreds, or even thousands, of pounds — had little confidence they'd enjoy the bottle they were about to buy. Different problems on the surface; the same root underneath.

02The idea

Develop one AI consumer experience that solves both problems.

Rather than tackle the two barriers separately, the move was to build a single AI-powered consumer experience that solved both at once: give the consumer a confident, personalised recommendation — and, in doing so, generate richly detailed consumer flavour-preference data, leveraging cutting-edge AI.

03The AI role

AI as the engine: recommendations that sharpen over time.

The AI was the engine behind the experience, built in partnership with FlavorPrint.1 It turned each person's flavour responses into a personalised, confidence-scored recommendation, and sharpened over time. At scale, the same engine generated insight that lifted marketing efficiency across the mix — campaign development, ranging, paid media and new product development.

04The approach

A gamified flavour experience — personal recommendations, global data at scale.

Consumers rated a series of engaging visual flavour stimuli — a game, not a survey — and the AI returned hyper-personalised recommendations that overcame the confidence barrier at the moment of purchase. The same interactions generated global consumer flavour-preference data at remarkable scale and quality, wired across three commercial journeys — on-trade (a bar QR scan straight to a recommendation), off-trade retail and direct e-commerce — with a CRM opt-in capturing it as first-party data, and performance loops feeding retargeting and continual model improvement.

05The outcome

Two barriers down: more confident buyers, and flavour data at scale.

Both barriers came down. On the consumer side, the personalised recommendation and its confidence ('% match') score reduced the perceived risk at the point of purchase — with up to a third of people who completed the experience going on to buy, and up to 60% of them choosing a brand they had never bought before.4 On the business side, the platform built the flavour-preference asset Diageo had never had, going on to profile more than a million consumers worldwide across 21 markets and 16 languages.35 The validation followed: Diageo acquired the underlying flavour-matching technology in 2022,3 and extended flavour profiling into further categories, including cocktails, in 2024.5

Sources & substantiation
  1. Diageo launched the 'What's Your Whisky' AI flavour selector on 28 November 2019, using AI and machine learning with FlavorPrint to analyse a consumer's flavour preferences and recommend a whisky; live at launch in nine European countries and six languages. BeverageDaily, 28 Nov 2019. beveragedaily.com
  2. The barriers the experience set out to overcome — a complex, saturated category, risk aversion driven by an expensive price point, and a dated category image — and the FlavorPrint build (18 single malts mapped to a 33-dimension model of 500+ flavour data points; 11 visual flavour questions). Campaign Tech Awards finalist entry. campaigntechawards.com
  3. 'What's Your Whisky' was rolled out across 21 markets and 16 languages and integrated across retail, e-commerce and direct-to-consumer channels; Diageo acquired Vivanda — owner of the FlavorPrint technology behind it — on 30 June 2022. Diageo press release. diageo.com
  4. In travel retail (Paris Orly), Diageo reported that more than 30% of consumers made a purchase after completing the 'What's Your Whisky' experience — over 60% of them a brand they had never bought before. TR Business, 2 Jan 2024. trbusiness.com
  5. 'What's Your Whisky' has been experienced by more than one million consumers; in 2024 Diageo extended FlavorPrint flavour profiling into further categories, including the launch of 'What's Your Cocktail'. Diageo press release, 31 May 2024. diageo.com

The full case study — the engine design, the three consumer journeys and the first-party data model — is available on request. Request the full case study →

Case study describes work led by Andy Parton in a senior role at Diageo plc. All brands and trademarks are the property of their respective owners; Parallax Advisory is independent and not affiliated with or endorsed by them. Outcomes reflect a specific engagement and are not a guarantee of future results.

FAQ

What is Diageo's 'What's Your Whisky'?

What's Your Whisky is an AI-powered flavour-discovery experience Diageo launched in 2019. It uses artificial intelligence and machine learning, built on FlavorPrint technology, to analyse a person's flavour preferences and recommend a whisky they're likely to enjoy. It has since scaled to 21 markets and 16 languages and been experienced by more than a million consumers.135

How does the What's Your Whisky AI recommendation work?

The consumer answers a short series of visual flavour questions. AI and machine learning map those answers across a multi-dimensional flavour model — 18 whiskies profiled against more than 500 flavour data points — and return a personalised, confidence-scored recommendation in a single exchange.12

What is FlavorPrint?

FlavorPrint is a flavour-matching AI that maps a person's taste preferences against a large sensory database to recommend products they're most likely to enjoy. It was developed by Vivanda (originally out of McCormick & Company) and acquired by Diageo in 2022. FlavorPrint is a registered US trademark of its respective owner.3

How is AI used in spirits and drinks marketing?

Beyond novelty, the durable uses are personalisation and recommendation, first-party data capture, and the insight that then sharpens ranging, media and product development. The value comes from integrating AI across the consumer journey so the data compounds over time — not from any single isolated tool. What's Your Whisky is one example of that integrated approach.

What results did What's Your Whisky achieve?

Publicly, it scaled to more than a million consumers across 21 markets and 16 languages. In travel retail, Diageo reported that over 30% of people who completed the experience made a purchase — more than 60% of them a brand new to them. The flavour-matching technology was later acquired by Diageo and extended into other categories.345

Did Diageo keep using the technology after What's Your Whisky?

Yes. Diageo acquired Vivanda — owner of FlavorPrint — in 2022, and in 2024 extended flavour profiling into cocktails with 'What's Your Cocktail'. The same technology also powers the Journey of Flavour experience at Johnnie Walker Princes Street in Edinburgh.35

Can I see the detail behind the headline figures?

Some additional detail can be shared under NDA — please request the full case study or start a conversation below.

← Index ← Previous: Vuse Flavour

What's next

Two ways forward.

The full case

The engine design, the three consumer journeys and the first-party data model — available on request.

Request the full case

A conversation

About taking AI from a single integration to a marketing posture. FIELD or DEPTH.

Start a conversation