Pınar Su Closes To Within Four Points Of Category Leadership In AI Answers
Second in category visibility. The four-point gap to the leader marks a measurable and closable range rather than a structural disadvantage.
AI visibility is now a measurable layer in Pınar Su ve İçecek's digital growth programme. Working with Webtures on SEO and Generative Engine Optimization (GEO), the brand treated its presence in organic search and in generative answer surfaces as a single problem. A baseline measurement, run over a category-specific question set, showed how often, in what position and from which sources the brand is cited across AI assistants. The findings are clear: when Pınar is mentioned it appears in the most prominent position and with the most positive tone in its category; the real gap is not the quality of mention but the breadth of coverage and whether the brand's own content is cited as the source. This is not a retrospective growth chart. It is a baseline diagnosis that fixes today's position in data.
Pınar Su ve İçecek sits inside one of the entity clusters generative engines most often merge: three separate publicly listed companies under a single umbrella brand. The programme started precisely here, from getting the brand recognised correctly by machines, and placed its AI-search visibility on measurable ground.
Why Pınar Su Needed SEO and GEO Support
Pınar Su's need arose from four structural challenges specific to the bottled-water category. Their common thread is that, despite strong brand awareness, the correct information could not be read by machines from the brand's own source. These four points were identified at the outset, and the brand partnered with Webtures to strengthen its digital visibility.
- Entity confusion: Pınar Su ve İçecek, Pınar Süt and Pınar Et are three separate publicly listed companies under the same umbrella brand. Add product brands such as Pınar Madran, Pınar Denge and Pınar Yaşam Pınarım, along with Yaşar Holding's ownership stake, and the result is an entity cluster that AI assistants routinely conflate. The answer to "which company is Pınar Su" had to be learnable from the brand's own content.
- Corporate data locked inside PDFs: The brand's most distinctive data — for example, having produced Türkiye's first bottled water, or its share of national water exports — lives inside annual-report PDFs with no HTML equivalent. Generative engines hit the same wall; this is a concrete AI visibility gap the brand can close on its own site.
- Local, dealer-dependent ordering: Demand for large-format water forms at the district level and is met through a dealer network exceeding a thousand outlets. Dealers using the brand name on their own sites creates cannibalisation in local results; the consistency of business data directly determines local visibility.
- Two separate regulatory regimes: Spring and drinking waters fall under the Regulation on Water Intended for Human Consumption, while natural mineral waters are subject to a separate regime. Every claim about water quality and mineral content must respect this distinction.
Goals Set for the Pınar Su Project
The project's goals centred on getting the brand recognised correctly as its own legal entity and making its AI-search visibility measurable. Specifically: associating Pınar Su ve İçecek accurately with its own product brands, moving corporate data out of PDFs into citable HTML passages, meeting the local intent in large-format water ordering without conflicting with the dealer network, and improving visibility across generative answer surfaces.
Goals were defined together with the regulatory frame from the start. Statements on water quality and mineral values were written to the limit permitted by the relevant regulations and the Regulation on Nutrition and Health Claims. Measurability was set as a separate goal: the brand's AI visibility would be measured regularly against a category-specific question set and reported alongside the competitor band.
How Webtures Approached SEO and GEO for Pınar Su
Webtures applied a strategy that combined technical SEO, content architecture and generative AI visibility. The premise is simple: there is no separate mechanism for appearing in generative answers; being indexable and snippet-eligible is the precondition for being cited on those surfaces. The strategy advanced across four workstreams.
Technical SEO and AI Accessibility
Crawlability was reviewed so that both search engines and AI systems could parse the content. The most decisive finding here concerned format: all verifiable corporate data was locked inside PDFs. Even an indexable PDF is far less suited to passage-level citation than HTML; topics such as the brand's history, export data and ownership structure needed an HTML equivalent. The division of roles between the corporate site, the ordering site and the umbrella-brand site was also clarified, because the measurement showed that the brand's tracked domain did not appear as a source in answers.
Content and Semantic Authority
Content was developed around the questions people actually put to search engines and AI assistants. In this category, questions fall into two clusters. The first is informational: the difference between mineral water and soda, the meaning of mineral values in a water analysis report, which regulation spring water falls under. The second is transactional: large-format water prices, subscriptions, the nearest dealer. On the same layer, the entity distinction was made explicit; the company's own product range, ownership structure and history were told independently of the separate dairy and meat companies.
GEO and AI Visibility
Work was carried out to strengthen the brand's presence across AI-assisted search experiences such as ChatGPT, Gemini and Perplexity. A category-specific question set was defined and generative answers were measured against it. Part of the set was deliberately built to test the entity distinction: whether an assistant confuses Pınar Su with Pınar Süt, and whether it links product brands to the correct company. Measurement was placed within a framework that tracks the brand's mention rate, its average position within the answer, its sentiment score and its source citation together.
Experience and Conversion
Page structures and user journeys were arranged so visitors could reach information and place orders easily. In large-format water ordering, intent is completed not in the content but in the app and over the phone, so the page's job is to shorten the handoff. The dealer-finding journey was treated as a direct extension of local visibility; inconsistent business data here corrupts not only the user experience but also the model that search engines and generative systems hold of the brand.
Measuring Pınar Su's Visibility in AI Search
Pınar Su's visibility in AI search was fixed as a baseline measurement over a category-specific question set. This is not a retrospective trend but a diagnosis that establishes today's position in data; it shows how often, in what position and from which sources the brand is cited in generative answers. The findings below form the reference point against which subsequent measurements, using the same question set, will be compared.
Measurement Framework and Question Set
The measurement was built on a question set reflecting the real intents of the bottled-water category. Questions cover topics such as glass-bottle water supply for restaurants and events, large-format water subscriptions for offices and homes, natural spring water certification, and the safety of glass versus PET bottles. Intent ran along three axes: Informational, Commercial Investigation and Local. Answers were monitored across ChatGPT, Gemini and Perplexity for the Turkish market. The aim of the framework was not to reduce everything to a single number, but to read mention rate, position, tone and source together.
At the question-set level the brand's coverage was strong: Pınar Su entered the answer in the large majority of the tracked category questions. The one clear topic where it fell outside coverage was the query focused on natural spring water certification; the brand did not appear in that answer at all. This gap is not random but directly content-related: when certification and source-identity information is not available in citable form on the brand's HTML, the model answers that question from other sources.
In-Category Visibility and Competitor Comparison
In in-category visibility, Pınar Su ranked second within the tracked brand band. Share of voice — the brand's mention share in generative answers — was measured at 18% for Pınar, while the band's leading competitor sits ahead at 22%; two brands follow at 6% and 0%. The gap is four points, meaning the lead is not distant but within a closeable range. The table below summarises the competitor comparison from the baseline measurement (competitor brands have been anonymised).
| Brand | Visibility (Share of Voice) | Total Mentions | Average Position | Sentiment Score |
|---|---|---|---|---|
| Category-leading competitor | 22% | 11 | 4.44 | 56 |
| Pınar Su ve İçecek | 18% | 9 | 2.33 | 59 |
| Third competitor | 6% | 3 | — | 51 |
| Fourth competitor | 0% | 0 | — | 51 |
Source: Brantial AI visibility measurement — baseline snapshot. Visibility shows the brand's mention share (share of voice) across the tracked generative answers. For average position, a lower value is better (the brand is listed more prominently in the answer). Sentiment score is on a 0–100 scale. Competitor brand names have been withheld while preserving comparison integrity.
The table shows how the raw "how many times mentioned" count alone can mislead. The leading competitor is mentioned more (11 mentions), but Pınar — despite being mentioned less (9 mentions) — is the best in the category on both average position and sentiment. In other words, Pınar's visibility gap is not an authority or perception problem; it stems from the brand entering fewer questions, that is, from the breadth of coverage.
Quality of Mention: Position and Sentiment
On quality of mention, Pınar Su leads the category. The brand's average position within the answer was measured at 2.33, meaning that in the answers where it is listed it mostly appears in the top places. By comparison, the leading competitor's average position is 4.44, further down. Recalling that a lower value is better on the position metric, Pınar holds the most visible position when it is mentioned. In some queries the position drops to 1; particularly on intents around functional and certified products, the brand is the first name mentioned.
Sentiment tells the same story in Pınar's favour. The brand's sentiment score of 59 is the highest in the category; the other tracked brands stay in the 56 and 51 band. Read together, these two metrics give a clear picture: Pınar Su is mentioned both prominently and positively in generative answers. The strategy's priority, therefore, is not to fix how the brand is mentioned but to get it mentioned in more questions and from its own source.
Source Citation and Content Gaps
The most critical finding of the baseline measurement is on the source side. Pınar Su is mentioned in generative answers, yet the brand's own domain is cited as a source in those answers almost never. Among the sources the model relies on, third-party wholesale and price-comparison sites and competitor brand domains stand out. This overlaps directly with the "corporate data locked inside PDFs" finding: the brand is known, the model mentions it, but it draws that information not from the brand's own HTML content but from second-hand sources.
Content gaps take concrete shape within this frame. The certification-focused query where the brand fell outside coverage and the absence of its own domain as a source point to the same root cause: verifiable corporate information not having been moved into citable HTML passages. Because the baseline measurement names these gaps, the impact of subsequent content work becomes measurable over the same question set.
What Was Carried Out in the Pınar Su Project
Webtures delivered work tailored to the brand's needs across technical SEO, content architecture, site structure, structured data and GEO. The main workstreams prioritised closing the coverage and source gaps the baseline measurement pointed to.
- Crawlability, indexation and page-experience audit
- Moving corporate data left in PDFs into citable HTML passages
- Clarifying company and product-brand relationships through structured data
- Role separation between the corporate, ordering and umbrella-brand sites
- Question-led content architecture with context-independent answer passages
- Content review compliant with the two separate water-regulation regimes
- Consistent business data and local-intent work for the dealer network
- Question-set monitoring across generative answer surfaces, with competitor-band tracking
What the Baseline Measurement Revealed
The programme was tracked not through a single metric but across three reinforcing visibility layers: organic search performance, local and dealer-based visibility, and citation in generative answers. The baseline measurement in hand gave a clear reference point for the generative-answer layer. This is not a growth chart; it is the zero point against which subsequent measurements will be compared.
Position Within the Category Band
Within the category band, Pınar Su ranks second at 18% visibility, four points behind the leader. This shows the brand is not invisible in generative search but, on the contrary, in a strong starting position. The source of the gap was also named by the measurement: the leader enters more questions. The measurable goal is therefore clear — increase the brand's coverage, that is, get the brand into the questions it currently falls outside of, from its own source.
From Being Mentioned to Being Mentioned Correctly
In a multi-brand structure, visibility is measured not by being mentioned but by being mentioned correctly. When Pınar Su is mentioned it appears in the most prominent position in its category (average position 2.33) and with the most positive tone (sentiment 59); the problem is not quality. The real gap concentrates in two areas: the brand's own domain not being cited as a source in answers, and its absence from certain questions altogether. Both gaps close when corporate data sits in machine-readable form on the brand's own source. Because the baseline measurement puts these gaps into numbers, the impact of subsequent work will be measured with the same framework rather than estimated.
The Outcome of the Pınar Su and Webtures Partnership
The SEO and GEO strategy Webtures applied addressed Pınar Su ve İçecek's organic search performance and its visibility in AI-assisted search experiences together. The baseline measurement showed the brand in a strong but incomplete position: prominent and positive when mentioned, yet with coverage and source share below where they should be because it is not fed by its own content. This picture also makes concrete why, in a multi-brand structure, visibility must be measured by being mentioned correctly rather than merely being mentioned. When corporate data sits in machine-readable form on the brand's own source, both coverage and source share will rise measurably; the next measurement round will show that progress over the same question set.
Frequently Asked Questions
How can AI visibility be reported from a single measurement
AI visibility can be reported as a baseline measurement even without historical data. In this approach, the brand's mention rate over a specific question set, its position within the answer, its sentiment score and its source citation are fixed in a single snapshot. The report makes no "increase" claim; it establishes today's position in data and sets the reference point against which subsequent measurements will be compared. In Pınar Su's case, this baseline clearly laid out the brand's place in the category band and the gaps to be closed.
What is the difference between share of voice and mention count
Share of voice shows the brand's visibility share in generative answers, while mention count shows the raw number of repetitions. The two do not always point the same way. Pınar Su, despite being mentioned less than the leading competitor, is the best in the category on average position and sentiment; in other words, it is mentioned less often but more favourably. Visibility should therefore be assessed not by a single number but by reading rate, position, tone and source together.
If a brand is mentioned in generative answers, why should its own site be cited
Even if a brand is mentioned, if the source the answer relies on is not the brand's own site, the accuracy and currency of the information are left to third parties. In the Pınar Su measurement, the model mentions the brand but cites third-party and competitor domains as sources. Having corporate data sit in citable form on the brand's own HTML both ensures the correct information is read from the brand's source and increases its visibility share.