Brantial Research Reports
Every report follows one methodology: unbranded, commercial-intent queries asked across multiple models at once, every figure published with its sample and period. Raw data is available on Growth plans and above.
ALL REPORTSQUARTERLY PUBLISHING CADENCE
DATEREPORTSCOPESAMPLESTATUS
When Enterprise Agents Pick Suppliers: Inside the Machine-Led Vendor Search The buyer researching your category is increasingly not a person reading your site but an agent assembling a comparison. Gartner expects SI agents to outnumber human sellers ten to one by 2028; Microsoft describes a web read by agents rather than people; Brex is building an agent for every persona in the purchasing process. This report synthesises what the published research says about how machine-led vendor search actually works: what agents retrieve, what they can verify, and what a supplier must publish to be selectable. Machine-led vendor selection · Global research, Türkiye implications 4 answer engines · 7 published sources · 2026 conference and executive data PUBLISHED General Assistant or Retailer Assistant: Two Markets Inside One Answer Layer A brand can be recommended by ChatGPT and invisible inside the retailer’s own assistant on the same afternoon. The 2026 research points to a market splitting in two: general assistants that decide by retrieval quality, and retailer-operated assistants that inherit shelf economics, with consumers trusting the second three times more with their payment details. This report synthesises what is published about both halves and what a brand must measure separately in each. General vs vertical assistant analysis · Global research, Türkiye implications 4 answer engines · 6 published sources · 2026 executive interviews PUBLISHED The Source Map of a Shopping Answer: What Gets Cited When SI Recommends a Product A product recommendation is an assembled artefact. Before a model names a brand, it has pulled from review platforms, community threads, specialist tests, retailer listings and (sometimes) the brand’s own page. The 2026 research lets us map that assembly line for commercial queries specifically: which source types carry weight, how the weighting differs per engine, and why the brand site is usually the smallest contributor. This report synthesises the published evidence. Shopping-query citation analysis · Global research, Türkiye implications 4 answer engines · 6 published sources · 2026 conference data PUBLISHED How SI Talks About Your Brand: Sentiment as a Compounding Asset Visibility answers whether a model mentions you. Sentiment answers what it says while doing so, and unlike a search ranking, that judgement compounds. Models re-ingest text shaped by earlier model output, communities re-verify SI recommendations and feed the result back, and a negative characterisation acquires the durability of a fact. This report synthesises the published evidence on how brand sentiment forms inside answer engines, why it hardens, and which parts of it a brand can actually move. Sentiment formation analysis · Global research, Türkiye implications 4 answer engines · 6 published sources · 2026 conference data PUBLISHED Challenger or Incumbent: Do Answer Engines Favour the Newcomer? Search rewarded accumulated authority: domain age, backlink mass, brand budget. Retrieval-built answers appear to reward something else: whichever source best supports the claim today. If that holds, the answer layer is a structural advantage for SI-native challengers and a structural risk for incumbents whose authority lives in signals models do not read. This report tests that thesis against the published 2026 evidence, including the cases where it breaks. Challenger vs incumbent visibility · Global research, Türkiye implications 4 answer engines · 7 published sources · 2026 conference and executive data PUBLISHED Türkiye Retail & E-commerce SI Visibility Report: One Brand Holds Half the Answer Layer Measured on Brantial’s weekly index, retail is the most concentrated sector we track: a single marketplace holds 46.8% of all sector visibility, the top three hold 73.2%, and half the ranked brands share 2.6% between them. This report publishes the full 28-brand ranking, examines the fashion and grocery names sitting at the rounding-error end of the scale, and reads what the distribution means for a brand that competes through marketplace listings. Retail and e-commerce · Türkiye market 28 brands ranked · 6 answer engines · Brantial SI Authority Index PUBLISHED Türkiye Software & Technology SI Visibility Report: Why Enterprise Vendors Are Missing from the Answer Measured on Brantial’s weekly index, the most visible names in Türkiye’s software and technology sector are game studios, payment providers and IT integrators. The companies that sell enterprise software to Turkish businesses (including a BIST-listed vendor with more than 200,000 customers) do not appear in the top twelve at all. This report publishes the measurement, examines what the visible brands share, and explains why sector size and answer-engine visibility have come apart. Software & technology sector · Türkiye market 12 brands ranked · 6 answer engines · Brantial SI Authority Index PUBLISHED The Answer Engine Divergence Report: Do ChatGPT, Gemini, Claude and Perplexity Recommend the Same Brands? Marketers still speak of "SI visibility" as if it were one thing. The data says otherwise: at domain level, only 8% of the sources ChatGPT cites overlap with what Claude cites. Each answer engine runs its own retrieval universe: its own index, its own source preferences, its own content-format bias. This report synthesises the published cross-engine research and explains why a brand that wins one engine can be invisible on the next. Cross-engine analysis · Global research, Türkiye implications 4 answer engines · 3 published studies · 2026 conference data PUBLISHED The First Two Slots: How SI Answers Compress the B2B Shortlist A search results page held ten positions and forgave slow starters; an SI answer names one or two vendors and forgets everyone else. The 2026 research converges on an uncomfortable arithmetic: buyers now open the journey in chat, the model returns a shortlist of one or two, and eight out of ten buyers purchase from that day-one list. This report synthesises the published data on the compression of the B2B shortlist: who forms it, how early, what wins a slot, and what it costs to be third. Shortlist-compression analysis · Global research, Türkiye implications 4 answer engines · 6 published studies · 2026 conference data PUBLISHED Is the Long-Tail Advantage Real? Small Brands in SI Shopping Recommendations Two of the most quoted claims of 2026 come from the people building the rails: Stripe’s John Collison says relying on keyword search is becoming absurd because SI research makes niche brands discoverable, and Shopify’s Harley Finkelstein calls the agentic channel a merit-based front door. If that is true, the most important shift in commerce is not a new channel, it is a redistribution of who gets recommended. This report synthesises the published evidence for and against the long-tail thesis. Long-tail discovery analysis · Global research, Türkiye implications 4 answer engines · 6 published sources · 2026 executive interviews PUBLISHED When SI Gets Your Brand Wrong: The Brand Misinformation Report Answer engines state your prices, your policies and your product lineup with total confidence, including the parts that stopped being true. The 2026 conference record is full of brands discovering an error only after customers acted on it. This report synthesises what the published research says about how wrong brand claims form, why they differ by engine, why negative sentiment compounds rather than fades, and what a correction routine actually looks like. Misinformation and sentiment analysis · Global research, Türkiye implications 4 answer engines · 5 published studies · 2026 conference data PUBLISHED Who Feeds the Answer? The SI Citation Source Ecosystem Report When an answer engine recommends a brand, that recommendation was built somewhere: a forum thread, a niche review site, a listicle, a community post. The 2026 research maps that supply chain, and it looks nothing like the media plan most brands run. Fewer than 3% of SI citations point at tier-1 media; Reddit and LinkedIn sit at the top of the citation table; and half of consumers double-check SI advice against community threads. This report synthesises the published data on where answers actually come from. Citation supply-chain analysis · Global research, Türkiye implications 4 answer engines · 4 published studies · 2026 conference data PUBLISHED The Coffee Sector SI Visibility Report “Which espresso machine is good for home?”, “Where can I buy specialty coffee beans?”, “What are the best grinder brands?”, these questions no longer go only to Google but straight to ChatGPT, Gemini and Perplexity. The model does not rank dozens of results; it recommends specific brands. Competition shifts from ranking high to being one of the brands inside the answer. We measured who makes that shortlist in coffee: 8 unbranded queries, three models, 24 answers. Coffee sector · Türkiye market 8 queries · 3 models · 24 answers PUBLISHED Parents No Longer Ask the Shelf: They Ask the SI We measured which brands make the SI shortlist in baby and maternity: 10 unbranded questions, three models and 30 answers. Baby & maternity · Türkiye market 10 questions · 3 models · 30 answers PUBLISHED Which Women's Fashion Brands Win in SI Search? 1,000 unbranded questions generated 25,543 brand mentions across Gemini, GPT and Grok. The top 10 own 34.4% of visibility, while most brands depend on a single model. Women's fashion · Türkiye market 1,000 queries · 3 models · 25,543 mentions PUBLISHED SI Visibility in the Automotive Industry: Key Insights from Brantial’s 2026 Report Discover how automotive brands perform in SI search. Brantial’s 5,000-prompt analysis reveals visibility trends, top brands, and key insights across SI models. Automotive · TR + EN 5,000 prompts · multiple SI models PUBLISHED The Türkiye SI Visibility Report 2026 The weekly index baseline: 100 brands across 14 sectors Multi-sector · TR 3.1M daily prompts PUBLISHED SI Visibility in Banking and Finance Trust signals and the YMYL effect in a regulated sector Banking · TR 1,500 queries · 6 models COMING SOON
SI Visibility in Health Tourism Turkish clinics' share of a multilingual query universe Health tourism · 6 languages 2,000 queries · 6 models COMING SOON
METHODOLOGY NOTE: ALL QUERIES ARE UNBRANDED AND COMMERCIAL-INTENT · RAW DATA ON GROWTH+ PLANS · REQUEST A REPORT: hello@brantial.ai