JavaScript Disabled: This site works best with JavaScript enabled. However, you can access the essential information below.
01 · At a glance
The engagement in one panel
- Client
- Trip Drinks
- Category
- Functional drinks
- Market
- United Kingdom
- Engagement
- 60-day AI search optimisation sprint
- Surfaces tracked
- Leading large language models, tracked with Share of Model
- Measurement
- Share of Model: share of voice, average position, mention rate
- Client lead
- Barnaby Whicher, eCommerce Lead, Trip Drinks
- Lmo7 lead
- Stephen Honight, Founder
- Primary metric
- Average position in AI recommendation lists
- Result
- 7th to 3rd average position, +33% AI-referred traffic (sources in Results below)
02 · The situation
Strong brand, weak placement in AI answers
Trip had the products and the brand awareness. What it did not have was a reliable place in the recommendation sets AI assistants produce when a shopper asks for a functional drink, a calming drink or an alternative to alcohol. Across the prompt set we tracked, Trip's average position sat at 7th.
That matters because AI recommendation lists are heavily top-weighted. A shopper reading an AI answer rarely gets past the third brand named. Outside the top three, visibility drops fast and the chance of being chosen drops with it. More mentions on their own would not fix that. The job was to move Trip up the list.
03 · How AI assistants decide who to recommend
Why average position is the metric that pays
When someone asks an AI assistant for a product recommendation, the model retrieves what it can find and trust about the brands in that category, then composes a short list. Where a brand lands depends on three things: whether the model can read the brand's own pages, whether sources it already trusts mention the brand in that context and whether the content answers the question being asked in language the model can lift cleanly.
Mention rate shows a brand is in the conversation. Average position shows whether it is winning it, and for a challenger brand that second number is the one that turns into sales. That is why it was the priority metric for this engagement.
04 · What Lmo7 did
A focused 60-day programme, tracked throughout
Baseline and measurement
We built a defined prompt set for the functional drinks category and tracked it across the leading models with Share of Model. Three metrics ran throughout: share of voice, average position and mention rate. Tracking ran throughout the sprint, so every change could be judged against the number it was meant to move.
Site signals
We improved the signals AI systems read first: page structure, entity clarity, structured data and the technical readability of the pages models retrieve most often. These are the fixes that make a page retrievable in the first place.
Priority content
We identified the Trip pages AI systems most commonly referenced and upgraded them so they answer category and product questions directly, in the terms shoppers use. Content was written for two readers at once: the person deciding what to buy and the model deciding what to recommend.
Citations and supporting references
We strengthened the third-party references and citations that models lean on when judging whether a brand is credible in its category. This is the slower lever and the one that holds a position once it is won.
05 · Results
What moved in 60 days
| Metric | Before | After | Source and window |
|---|
| Average position in AI recommendations, functional drinks | 7th | 3rd | Share of Model tracking across the defined prompt set and tracked models, 60-day engagement, results published 27 February 2026. |
|---|
| AI-referred site traffic | Baseline | +33% | Trip Drinks site analytics, direct referrals from AI assistants, same 60-day window. |
|---|
Moving into the top three is the result that counts, because that is where shoppers stop reading. The traffic uplift shows the placement change translated into shoppers clicking through to Trip.
"Lmo7 have provided us real clarity on our AI search visibility across the major models and a practical plan to improve it. As eCommerce Lead at Trip Drinks I've been on the forefront of optimising for AI search, focused on actually moving the needle."
Barnaby Whicher, eCommerce Lead, Trip Drinks. LinkedIn recommendation.
06 · What this means for challenger brands
The playbook, and where it goes next
Trip's result came from a repeatable sequence. Measure with Share of Model so you know where you stand. Prioritise average position over raw mentions. Fix the signals and content that decide retrieval first, because they move fastest. Then build the citations and authority that keep you in the top three, because that is the part competitors cannot copy quickly.
The honest caveat is that the slow lever is slow. Site signals and content upgrades produce the early movement. Holding a top-three position depends on third-party authority, which takes sustained work well beyond a 60-day sprint. The next step for any brand at this stage is to keep tracking and keep building the citations that hold the place.
The Lmo7 Agency - AI Search and Ads for Consumer Brands | UK-based | Serving Global Brands
For the full interactive experience, please enable JavaScript in your browser.