Q1. Consumers are increasingly turning to AI assistants instead of traditional search engines for recommendations. How is this shift changing the way brands are discovered online?
Earlier, people used to go to Google, Bing and other search engines when they had a question. Today, they ask ChatGPT, Gemini, Claude and Perplexity. Even when they do go to Google and type in a query, they often end up reading the AI summary at the top rather than going through ten links.
Though traditional search will not vanish anytime soon, for most people, the discovery moment now lives inside a single conversation. Which means shortlisting has already happened before the person even sees a name. Digiday’s reporting on the last holiday shopping season captured this shift precisely: people are starting their research, comparison and purchase journey inside AI chatbots such as ChatGPT, Gemini and Perplexity. It has also changed the shape of the query itself, from short keywords to longer, conversational asks with personal context built in.
AI can now go a step further and complete the purchase itself. That’s the next evolution of agentic commerce: the shift from AI recommending a product to AI actually buying it on someone’s behalf, using stored preferences and payment details to complete the transaction with minimal human intervention.
What this means practically is that a brand isn’t just competing for rank anymore; it’s competing to be part of the answer. If a brand’s product information isn’t structured well enough for an AI system to read, verify and recommend with confidence, that brand can simply be left out of the recommendation, no matter how strong its traditional SEO is. Furthermore, AI systems rely on what trustworthy third parties say about a brand: editorial coverage, independent reviews and credible mentions across the web, rather than paid placements or advertorials that read as self-interested.
The same shift is already beginning to extend into transactions. In India, NPCI’s work on enabling AI-based commerce points towards a future where AI won’t just recommend a product but could complete the transaction too. Once an AI assistant is connected to live inventory systems, it could recommend the exact product that fits a customer’s needs, say, the right tyre for a specific vehicle, confirm stock at a nearby dealer, compare prices in real time and then place the order, all within the same conversation. That’s a meaningfully bigger shift than better discovery, and it’s an area we at FlowBlinq are building towards directly.
Source: Digiday
Q2. SEO has been the cornerstone of digital marketing for years. With AI-generated answers becoming more common, how should marketers think about the transition from SEO to AI visibility or Generative Engine Optimization (GEO)?
SEO isn’t going away. What changes is the job each discipline is doing. SEO was built to help a brand rank in search results and earn a click. GEO is about making sure that brand is understood, cited and recommended when the answer is generated directly by an AI system.
The shift is already significant. McKinsey’s research from last October found that half of consumers are now using AI-powered search, a behaviour it estimates could influence roughly $750 billion of US revenue by 2028. Gartner has gone further, projecting that traditional organic search traffic could decline by more than 50% by 2028 as this behaviour grows.
So this isn’t really a question of moving from SEO to GEO. Marketers need to run both in parallel, but understand that they solve different parts of the discovery journey. SEO still helps earn the visit; GEO helps earn the recommendation when there may be no visit at all.
That means the fundamentals of SEO still matter, but they need to be complemented by things AI systems can interpret and verify: clear content structure, consistent entity information, machine-readable product data and credible third-party signals. The brands that treat SEO and GEO as competing disciplines risk optimising for one discovery environment while becoming less visible in the other.
Q3. Agentic Commerce is emerging as the next evolution of online shopping. What does it mean for brands, and how is it different from the e-commerce ecosystem we know today?
Traditional e-commerce assumes a human does every step: searching, comparing tabs, adding to cart and entering card details. Agentic commerce removes the human from several of those steps and hands them to an AI agent acting on the person’s behalf, from discovery through to a completed transaction.
The clearest example is what OpenAI and Stripe built last September with Instant Checkout inside ChatGPT, powered by the Agentic Commerce Protocol, an open standard the two companies co-developed so merchants can plug into agent-led buying without rebuilding their backend. Salesforce joined the same protocol a few weeks later. Their own study found that 48% of shoppers who already use AI for shopping would be comfortable letting an AI agent complete a purchase for them. That’s a strong indicator of where commerce is headed.
For a brand, this changes what they are actually optimising for. In e-commerce, you’re designing a page to convince a person; in agentic commerce, you’re also structuring your product data, pricing, availability and policies so an autonomous agent can read them accurately, trust them and act on them without a human double-checking every detail. That’s an infrastructure problem as much as a marketing one, and most brands haven’t started treating it that way yet.
Source: Stripe, Salesforce
Q4. Many marketers still rely on website traffic and click-through rates to measure digital success. In an AI-first world, what new metrics should brands be paying attention to?
Click-through rate assumes a person needs to click to see your brand, or your product. However, like we discussed, this is changing when everything is happening inside a single chat interface. When an AI system answers a question directly, someone can choose a product without a single visit ever being logged. Siteimprove’s analysis puts a number on it: when an AI overview appears on a search results page, organic click-through drops by 61%, and none of that gets picked up by a rank tracker, because those tools were never built to register a citation.
Three metrics matter now: citation rate, how often your pages get linked as a source in an AI answer; mention rate, how often you’re named without a link; and share of voice, your slice of citations against named competitors on the queries that matter to your category.
None of these come out of your existing analytics stack. You have to build a monitoring layer that actually queries ChatGPT, Gemini, Perplexity and Claude the way your customers do, and tracks how your brand shows up over time, which is exactly why we built AI Citations at FlowBlinq, to give marketing teams a working view of this instead of guesswork. That’s a new muscle for most marketing teams, and it’s the one I think matters most over the next two years.
Source: Siteimprove, Data Mania
Q5. As AI agents begin recommending products, booking services, and even completing transactions on behalf of consumers, how should brands prepare for a future where customers may never visit their website?
The honest starting point is accepting this is already happening at a meaningful scale, not some distant scenario. Adobe Analytics reported AI driven traffic to US retail sites growing nearly 4,700% year over year in 2025, and Amazon saw AI shopping assistant traffic jump 3,300% on Prime Day alone.
Preparing for it means treating your product catalogue, pricing, policies and availability as an API for AI agents to consume, not just a webpage for humans to browse. That includes clean structured data, accurate product codes, and no accidental blocks on AI crawlers. When we audited Indian merchant websites at FlowBlinq, 47% were unknowingly blocking ChatGPT from reading their site, usually because a security plugin had quietly added the restriction, and 91% had no proper guide telling AI systems what they sell or how the catalogue is organised.
Brands also need a plan for the moment an AI, not a human, is confirming a transaction. That means checkout, returns and customer service logic all need to be legible to a machine acting in good faith on a customer’s behalf. The website doesn’t disappear, but it stops being the only front door, and for a growing share of customers, it may just be a source of information for their AI shopping agents.
Source: Adobe
Q6. What are some of the biggest misconceptions brands have about being discoverable by AI platforms? Is strong SEO alone enough anymore?
The biggest misconception is that good SEO automatically carries over to AI visibility. It doesn’t, because the two systems are evaluating fundamentally different things. Search engines rank pages; AI systems select and cite sources they trust enough to quote directly, and that trust comes from a different set of signals altogether, how consistently your brand identity appears across the web, whether your content can actually be extracted and quoted cleanly, and whether independent sources corroborate what you say about yourself.
The second misconception is treating this as purely a content problem. A lot of it is technical. Our own research at FlowBlinq found 62% of merchant websites had product descriptions too thin for an AI to describe or recommend accurately, and 54% were missing basic product codes AI systems need to match and verify listings before they’ll cite them with confidence.
So no, strong SEO alone isn’t enough anymore. It gets you found by a crawler; it doesn’t guarantee you get cited by a model deciding what to tell a customer in the next three seconds. Those are related disciplines, not the same one, and brands that keep treating them as identical will keep wondering why their rankings look fine while their AI visibility doesn’t.
[Source]
Q7. In your view, which industries are likely to see the earliest impact of Agentic Commerce, and what lessons can other sectors learn from them?
Retail is out in front, which isn’t surprising given how transactional and structured the category already is. Grand View Research’s latest sizing puts retail and e-commerce as the dominant application for agentic commerce in 2025, with the overall global market valued at $5.7 billion last year and projected to grow to $65.5 billion by 2033.
Travel is close behind, and for good reason. Travel and hospitality suit agentic commerce naturally, agents can manage planning, pricing changes and multi-step bookings, all things a human normally juggles across several tabs. Financial services and payments are moving too, with players like Visa and Mastercard building agent specific payment rails so transactions initiated by AI can be verified and trusted the same way a human initiated one is.
The lesson for every other sector is that agentic commerce rewards categories with structured, comparable data, prices, availability, specifications, cancellation terms. If your industry’s information still lives mostly in PDFs, phone calls or ambiguous prose, that’s the work to start on now, well before agents become the default way your customers transact.
Source: Grand View Research, Fintech Futures
Q8. Trust has always influenced consumer decisions. In the age of AI, what factors determine whether a brand is cited or recommended by platforms like ChatGPT, Gemini, or Perplexity?
It comes down to three things: whether the AI can clearly identify who you are, whether it can easily extract facts from your content, and whether other independent sources back up what you claim about yourself. Entity consistency across your own site and third party listings, content structured for direct extraction rather than buried in long unstructured paragraphs, and outside coverage that confirms your own claims, that’s the trust ecosystem an AI is really evaluating.
The technical layer underneath that is structured data. A large-scale analysis of ten thousand queries across the major AI engines found FAQPage schema was the single strongest predictor of getting cited, with a correlation of 0.61, ahead of domain authority and content freshness. Organisation schema matters just as much, because it’s usually the first thing an AI system checks to establish whether it’s even dealing with a known, verifiable entity. But schema alone doesn’t build trust, it just makes trust legible to a machine. The trust itself still has to be earned through PR: independent press coverage, analyst mentions, and third-party reviews are what AI systems lean on to corroborate a brand’s own claims, and brands that have under-invested in PR in favour of pure content marketing are finding that gap show up directly in their AI citation rates.
The through line across all of this is that AI systems trust what they can verify quickly and confidently. A confident, well structured claim beats a vague, persuasive one every time, which is a genuinely different game from writing copy meant to persuade a human reader.
Source: Pixis, Authority Tech Globe Runner
Q9. For CMOs planning their digital strategy over the next 12–24 months, what capabilities should they prioritise to ensure their brands remain competitive in an AI-driven discovery landscape?
Data quality has to come first, unglamorous as that sounds. Adobe’s 2026 AI and Digital Trends report found 78% of CMOs cite data integration and quality as the single biggest barrier to adopting agentic AI properly, ahead of budget or talent. You can’t be recommended by a system that can’t trust or parse your data, so this is the foundation everything else sits on.
Second, CMOs need to build measurement muscle for a channel their existing tools weren’t designed to see, standing up citation and share of voice tracking as seriously as they track paid media performance today.
Third, and this is where budgets are already moving, is genuine investment rather than experimentation. Gartner’s 2026 CMO Spend Survey found CMOs are now allocating 15.3% of their marketing budget to AI, yet only 30% describe their organisation as actually ready to scale that investment. Closing that readiness gap, through data infrastructure, upskilled teams and clear ownership of AI visibility as a workstream, is what separates brands that compound their advantage from the ones still running pilots two years from now
Fourth, and often overlooked, is PR. Since AI systems weigh earned, third-party coverage more heavily than brand-owned content when deciding what to trust, a strong PR programme is no longer just a reputation function, it’s a direct input into AI visibility, and CMOs should be resourcing it accordingly.
Q10. Looking ahead, how do you envision the customer journey evolving over the next five years? Will AI become the primary gateway between consumers and brands, and what does that mean for the future of marketing?
I think primary gateway is the right way to frame it, but it won’t happen evenly. McKinsey’s estimate that AI powered search could influence $750 billion of US revenueTwo numbers make the direction clear. McKinsey estimates AI-powered search could influence roughly $750 billion of US revenue by 2028, which shows how much of the customer journey is expected to move into these systems within a few years, not some distant future. Grand View Research separately projects the agentic commerce market growing from $5.7 billion in 2025 to $65.5 billion by 2033. Read together, both point to the same underlying pattern: budgets, infrastructure and consumer habits are all moving in the same direction, toward AI systems sitting between brands and the people who buy from them.
For marketing, the shift changes who the audience actually is. It has long been built to persuade a person standing at a decision point. Increasingly, part of that audience is an AI system making a recommendation on that person’s behalf. That system doesn’t respond to the same cues a person does. It responds to verifiable, structured, consistent information.
That doesn’t make brand building or creativity less important, if anything it raises the bar, because a distinctive, trustworthy brand is what an AI system reaches for when it has to choose between several technically similar options. The marketers who do well over the next five years will be the ones who can hold both truths at once, build a brand a human wants to choose, and structure the data so a machine is willing to recommend it.
Source: McKinsey, Grand View Research
