It used to be enough to show up on the first page of Google. Now brands need to appear in the AI answer itself. As consumers increasingly turn to ChatGPT, Gemini, and AI-powered search for recommendations, spirits brands must adapt how they show up online.
In Part 2 of this Park Street University series on AI for spirits marketing, Antonia Fattizzi, Founder and President of Cork and Tin, continues her conversation with Justin Noland, VP of Digital Experience for Treasury Wine Estates, diving into practical execution.
Writing Effective Prompts
Context Is Everything: Noland emphasizes that strong prompts begin with context. Tell the AI what role it should play (social media manager, insights analyst), explain the entire problem, and specify the desired output format. His approach: treat AI like “an incredibly brilliant intern” that needs clear direction.
Ask the AI for Help: If unsure how to write a prompt, simply ask the AI what prompt you should write to accomplish your goal. Or request that it ask five clarifying questions to help craft the best possible prompt.
Start Fresh When Needed: AI tries to statistically deliver the most likely outcome based on conversation history. If threads get long or topics shift dramatically, start a new conversation to avoid confusion and maintain focus.
Maintaining Brand Consistency
Noland recommends creating a single brand document outlining brand story, tone of voice, content preferences, and topics to avoid. Attach this document to AI sessions when creating brand content to ensure consistency across all outputs.
Alternatively, direct the AI to pull information only from your website URL, ensuring it references your existing brand stories rather than external sources.
Optimizing Your Website for AI Discovery
With AI search volume increasing dramatically and poised to overtake traditional search, showing up in AI answers has become critical. Noland outlines two practical changes:
- Structure Your Data: AI engines view your website as the authoritative source for information about your brand and products. Ensure information is structured so AI can ingest and offer it back to consumers:
- Place clear definitions at the top of pages (who you are, what you do, what the product is)
- Add TLDR sections with five key bullet points about your brand history or story
- Give AI the most important information upfront
- Build External Validation: Unlike traditional SEO’s focus on backlinks, AI looks at what people are actually saying about brands. This includes:
- Critics and high-authority publications
- Social media mentions (which are now indexed)
- Volume and consistency of messaging across platforms
The more brands can get their story into media and journalist coverage, the better positioned they are for AI discovery.
The Consistency Factor
Fattizzi emphasizes that brands should maintain consistent messaging across websites, social media, and sell sheets. If a brand wants to be known as “the most refreshing” product, that language should appear consistently everywhere—on the website, in influencer campaign briefs, and in pitches to journalists.
This consistency helps AI connect unstructured content across platforms and increases the likelihood of appearing when consumers ask questions like “what’s the most refreshing beverage for a summer picnic?”
Common Mistakes to Avoid
Never Skip the Human Review: Noland’s top warning is to always keep a human in the loop. AI makes mistakes, and copying and pasting its output without review puts your brand reputation on the line, not the AI’s.
Pace Your Learning: These recommendations work in early 2026, but AI is changing at a dizzying pace. Brands should stay abreast of changes at least quarterly, if not monthly.
Support AI Adoption: For leaders, being supportive when staff members use AI for tasks like turning reports into executive briefs is crucial. As Noland notes, some tasks simply aren’t worth human time anymore when AI can handle them adequately.
Watch the full conversation above to hear Noland’s detailed examples on prompt engineering, website optimization, and building brands for the AI era.
AI for Spirits Marketing Pt. II: Putting AI Into Practice for Small Brands Transcript
Antonia Fattizzi & Justin Noland (0:02)
It is absolutely important to ensure that you are showing up in that answer. It used to be that if you showed up on the first page of Google, you were doing really well. Now it can’t just be the first page, it has to be the AI answer itself.
Antonia Fattizzi & Justin Noland (0:18)
Welcome back. I’m Antonia Fattizzi, founder of Cork and Tin, here with Justin Noland, VP of Digital Experience for Treasury Americas. This is part two of our video series on the intersection of AI and spirit brands, and today we’re going to be talking about the bridge from strategy to execution.
I think it makes sense to start from the beginning, where many people can get a bit stuck. What makes a good, strong, effective prompt, Justin? And equally, if you feel like you’re not getting what you need out of AI, what’s the best thing to do in that moment?
A good, strong, effective prompt begins with context. Context, context, context. The more information you can provide the AI about what you’re trying to accomplish, the better off you’re going to be. Start by literally telling the AI what you want it to act like. Do you want it to be a social media manager? An insights manager? What are you trying to have it do for you?
Always think of AI as an incredibly brilliant intern. They’re smart, they know an awful lot, and have access to more information than most, but they’re still an intern. So explain to them exactly how you want things to happen. Who do you want them to be? What do you want them to be like? Then explain the entire problem, exactly what you’re looking to get, and what you want the end output to be. “My end output needs to be a seven-paragraph blog.” “My end output needs to be a bullet-point list of the ten best takeaways from this material.” “I want you to act as an executive at my company reviewing my report and telling me the ten questions most likely to be asked by my board members or investors, so I can be more prepared.”
A great prompt is all about context, and sometimes it involves going back and forth. One of my favorite tricks: if you don’t know what kind of prompt to write, remember that AI works in natural language, natural language processing. You don’t have to be technical. If you don’t know what prompt to write, ask the AI: “What prompt should I write if I want to accomplish…” and then just write a paragraph in your own words describing what you’re trying to do. The AI will help you craft the prompt itself.
Antonia Fattizzi & Justin Noland (3:02)
Or, if you get to a point where you feel like you’re almost there but not quite, tell the AI: “Please ask me five more questions about what I’m trying to accomplish so that you can help me write the best prompt possible.” It will work through that with you and say, “Okay, here are all the context points I need to really deliver the output you’re looking for.”
I know that’s a lot to start with, but if you return to the idea of context the more you can tell the AI about what you want to accomplish and what you expect from it, the better the output will be. And if you find the output isn’t working for you, remember: it takes a lot of back and forth. It’s not necessarily going to give you everything you want the first time. That’s really how the pros work with AI, it’s a continuous conversation, a continual refinement of the language you’re using. Maybe you need to change a word because the AI’s definition of “luxury” differs from yours. Adapt the language and keep adjusting.
One of my favorite ways to think about it: AI is trying to statistically give you the most likely outcome you’re looking for. I mean that in a true mathematical sense, it’s all ones and zeros. AI is not a truly creative brain; it is looking at the statistics of what output is most likely to be what this person wants. And if you have a very long thread covering lots of different topics over several days, it can lose context. It might start pulling in aspects of your earlier conversations and mixing them into the current task, effectively thinking, “I don’t really know what this person wants. We were just talking about a spaghetti dinner and now they need board updates.” When that happens, start a new thread. Take all the previous context away and start completely fresh.
That really makes a lot of sense, and thanks for bringing the data points behind it into the conversation.
Antonia Fattizzi & Justin Noland (5:17)
We talked in the last video about not losing the essence of a brand when working with AI in a marketing context. So what recommendations would you have for incorporating a brand strategy directly into a prompt?
Definitely work it directly into the prompt. I always find it helpful to have a single document that outlines my brand, my brand story, my brand tone of voice, the kinds of content we like to create, the things we want to talk about, and the things we don’t. A pros and cons list of sorts. If I have that document sitting on my laptop, every time I start a new AI session I’m asking for brand content, brainstorming new ideas, and help with a sales sheet for a retailer. I can always attach that brand document so that every session starts with the same foundational information about the brand. That keeps things much more consistent and ensures the AI is delivering the right voice with the right stories, time after time.
You can also direct the AI to pull information from your website, so it’s not drawing from other online sources. If you have a wealth of brand stories on your own site, use the URL and say: “Only take information from here and help me pull together a sales sheet for this retailer.” There are lots of ways to do it, but the key is giving the AI clear direction about what your brand is across the board.
What I’m hearing in that is how much consistency plays a part in all of this. And that’s something I talk to my clients about constantly, their website should not look markedly different from their social media, which should not sound terribly different from what they’re putting on sell sheets. They all have their respective audiences and the story naturally has nuances depending on the application, but we shouldn’t be talking apples, oranges, and pears. That consistent jumping-off point from the very first iteration is really critical.
Antonia Fattizzi & Justin Noland (7:52)
Speaking of social media, I see a lot of smaller brands focusing almost entirely on social media as far as their budgets are concerned. But consumers are increasingly asking AI tools for recommendations. So if I own a small distillery, what are one or two practical changes I should make to my website to show up more clearly in AI responses?
We’re definitely seeing AI-driven search volume increase dramatically, and at some point it will overtake traditional search volume. Part of that is people using tools like ChatGPT or Gemini. The other part is that platforms like Google are incorporating AI directly into the responses they give, even in a regular search. So it is absolutely important to ensure that you are showing up in that AI answer — not just on the first page of results, but in the AI answer itself.
In terms of practical website changes: first and foremost, AI will largely treat your website as the strongest authoritative source for questions about your brand and your products. Something as specific as “What’s the ABV of this product?” — AI is going to look to your website because you’re the ones most likely to have it right. That’s a big component.
But you also need to make sure that if AI is going to treat you as the authority on your brand, you’re presenting content in a way it can actually ingest. Data structuring is the technical term for it — ensuring that the information on your product pages and across every aspect of your website is organized in a way that AI can read, understand, and relay back to consumers. Practically speaking, that means a clear definition at the top of the page: who you are, what you do, what the product is. If you have a long history or a rich story, think of it like a “TL;DR” — too long, didn’t read — section right at the top. Give the five key bullet points. If an AI doesn’t get to anything else on your page, what do you want it to take away?
Those are really the two most important things: structured data and front-loaded information. From there, AI engines don’t evaluate authority the same way traditional search engines do — it’s not about backlinks. It’s about what people are actually saying about you: what critics and high-authority sources are saying, and how much of it is out there. But it goes beyond just critics — social media is indexed too. It’s almost more democratized than traditional SEO. So think carefully about ensuring that all the voices out there talking about your brand are saying things you’d be happy with.
So it sounds like if a brand receives a rating, a write-up in a magazine, or gets listed in a roundup, they should be getting that onto their website just as quickly as they’d be posting it to their Instagram stories?
Antonia Fattizzi & Justin Noland (11:56)
Yes, as much as possible. This is almost a renaissance of public relations. The more you can get your brand out there with journalists and authoritative figures who can place it in media, the better off you’ll be.
You made a really good point about ongoing website maintenance and continued updates. The more consistent you can be in how you talk about your brand and your products, the more likely it is that AI will pick up on that consistency. If you want to be known as the most refreshing product out there, continually using that word in connection with your brand name is really important — not just on your own website, but in your influencer campaign briefs, when you’re pitching a journalist, in every touchpoint. You want that language being picked up and reinforced across the board. That is how AI pulls together all of those unstructured paragraphs of content from across the web and synthesizes them — so that when somebody asks an AI “What’s the most refreshing beverage I can bring to a summer picnic?”, you’ve got a much better shot of being the answer.
And I think it’s important to note that these are strategies that work today, in early 2026. But this is a space that is changing at a dizzying pace. I honestly don’t know how anyone stays fully across all of it, but I think it makes sense to recommend staying abreast of these changes at least every quarter, if not every month.
Antonia Fattizzi & Justin Noland (13:45 )
In your role at Treasury, you’ve structured AI education internally in phases for your staff. But for a founder team of one or two people, how should they pace themselves when it comes to upskilling?
First and foremost, just be supportive. Whether you’re a leader at a large organization or a small one, be supportive of your staff not just learning AI but using it on a day-to-day basis. There’s a tendency to treat it like a red flag — “Oh, they used AI for that” — as if it’s a bad thing, and that’s simply not true anymore. We’ve had situations on our own teams where someone says, “I have this huge report and I need an executive brief,” and leaders are now saying, “We don’t spend time on that anymore — an AI can handle it for us. It does it well enough, and it gets us where we need to be.” Being supportive of your staff using AI, learning AI, and coming up with new ways to utilize it is huge. It really matters.
That’s fantastic. On the flip side, what’s one common mistake smaller teams should avoid when they start experimenting with AI?
The first thing I always say: don’t remove the human from the loop. AI makes mistakes — know that going in. Don’t just copy and paste AI output and put it out into the world, because it’s not the AI’s reputation on the line. It’s still yours. It’s still your brand, your company, your personal reputation even within an organization. Always double-check it, make sure you’re getting what you want out of it, and don’t rely on it to be 100% accurate all the time.
Thanks so much, Justin. If you’re building a small spirits brand today, you’re well aware that you’re competing with companies that have larger teams and larger budgets. AI won’t replace strategy or creativity, but it can give smaller teams real leverage. Subscribe to Park Street University on YouTube for more insights from expert advisers across the industry. Thank you.