Milena Traikovich stands at the forefront of the shift from traditional search engines to Answer Engine Optimization (AEO). As an expert in B2B demand generation and performance optimization, she has spent the last year helping organizations navigate a landscape where buyers no longer just search for keywords, but engage in deep, contextual dialogues with artificial intelligence. Her work focuses on the intersection of data-driven lead generation and the evolving psychology of the modern buyer, specifically how brands can remain visible when the traditional rules of SEO no longer apply.
This discussion explores the radical shift in buyer behavior as AI chatbots become the primary starting point for B2B research. We delve into why traditional keyword tools are failing to capture the rich context of modern 60-word prompts, the critical importance of unbranded “how-to” content, and the high cost of keeping pricing behind “contact sales” buttons. Traikovich also breaks down the methodology for building a “Prompt Universe” and why capturing presence across hundreds of queries is more vital than ranking for a handful of keywords.
Modern AI prompts often average sixty words and include specific details like team size and budget that traditional keywords simply ignore. How does this shift in the way buyers phrase their needs change the core strategy for a B2B marketing team?
The shift is stark because we are moving from “searching” to “briefing.” When a buyer types a 3.4-word query into Google, they are looking for a door to walk through; when they give a 60-word prompt to ChatGPT, they are giving a set of requirements to a consultant. This means our strategy can no longer be about winning a specific high-volume keyword like “marketing automation.” Instead, we have to feed the AI the “context” it craves—details about team sizes, specific budget constraints, and integration deadlines. In my experience, if your content doesn’t explicitly mention that your software is optimized for teams of 50 or fits within a $10,000 monthly budget, the AI might filter you out before the buyer even sees your name. It’s a move toward radical specificity that many B2B teams find uncomfortable because they’ve been taught to be as broad as possible to capture the most traffic.
You’ve mentioned that unbranded questions are a massive blind spot for most brands, citing a test where ChatGPT chose HubSpot as the default solution for a general workflow query. Why are most AI visibility reports failing to catch these moments?
Most visibility reports are essentially just SEO tools with a new coat of paint, tracking a few dozen keywords rewritten as questions. They miss the “how-to” queries that don’t mention a brand because they aren’t looking for them, yet these are exactly the places where AI engines like ChatGPT make executive decisions for the buyer. In that specific lead nurturing test, ChatGPT didn’t just find an answer; it essentially became a HubSpot advocate because HubSpot’s own help guides and YouTube tutorials were the only ones providing clear, structured instructions that the AI could parse. Every other vendor was invisible because their content was either gated or too focused on “why” rather than the granular “how.” This blind spot is becoming incredibly expensive every quarter as more buyers rely on these unbranded paths to find their shortlists.
With 51% of B2B software buyers now starting their research with an AI chatbot—a significant jump from 29% just a year ago—what does this tell us about the trust buyers are placing in these platforms?
It tells us that buyers are exhausted by the “traditional” search experience, which has become a minefield of ads and SEO-optimized fluff. The jump to 51% shows a preference for the efficiency of a synthesized answer over a list of links that require manual vetting. There is a certain emotional relief for a buyer when an AI returns a clean, comparative table of six platforms instead of making them open twenty different tabs. However, the most telling statistic is that 69% of these buyers said the chatbot surfaced information that led them to a different vendor than the one they initially expected. This means the AI isn’t just a research tool; it’s a persuasion engine. Buyers trust the AI to do the heavy lifting of comparison, and if your brand isn’t part of that synthesized narrative, you aren’t just losing a click—you’re losing the chance to even be considered.
When an AI engine generates a comparison table for a buyer, it often pulls from dozens of different sources simultaneously. How can a vendor ensure their product is represented accurately in these high-stakes consideration moments?
Representation in those tables is won or lost in the technical clarity and accessibility of your public-facing data. When ChatGPT built that CMS comparison table, it didn’t just guess; it read 24 different sites, including review platforms like G2 and PeerSpot, as well as the vendors’ own technical documentation. To ensure accuracy, you need to provide the AI with “structured reality”—plain text pages that detail your developer flexibility, governance features, and cost structures. If your site is full of vague marketing speak like “world-class solutions” instead of “supports 5,000 concurrent users with 99.9% uptime,” the AI will either skip you or hallucinate a generic answer. I always tell my clients that the AI is like a very literal researcher; it needs hard numbers and clearly defined features to build those table rows, otherwise, it will pull that data from a third-party source that might not be as favorable to you.
The data suggests that 69% of buyers want to validate what an AI tells them with a sales rep, yet many vendors still hide their pricing. How damaging is this “contact sales” wall in an AI-driven research environment?
It is potentially fatal for a lead in the decision stage. When a buyer asks, “How much does AI customer support software cost?” and ChatGPT runs searches on Intercom, Zendesk, and Gorgias, it is looking for specific price bands. If your pricing is hidden behind a button, the assistant has nothing to quote, and you are effectively excluded from the “cost” part of the conversation. The AI might provide a range from free to over $10,000 based on your competitors, and you’ll be left out of the buyer’s budget calculation entirely. We are seeing that transparency is no longer just a sales philosophy; it’s a technical requirement for visibility. If the AI can’t find your contract terms or plan limits in plain text, it will simply direct the buyer toward the competitors who make that information easy to ingest.
If a marketing team discovers they are invisible in these chats, where should they start their content investment—is it explainers, comparison pages, or help documentation?
The starting point depends entirely on your market position and your most pressing business priority. If you are a challenger brand in a crowded category, you need to focus on awareness questions—those problem-first “explainers” and category guides that influence a buyer before they even have a shortlist. However, if you are an established vendor, your biggest risk is being swapped out during the comparison phase, which means you need to invest heavily in comparison pages and ensuring your presence on review sites is robust. For those focused on retention, the post-purchase documentation is the gold mine. As we saw with the lead nurturing example, help content is often the most powerful “pre-purchase” content because it proves the tool actually works. One vendor’s documentation can end up answering for an entire category if it’s the only high-quality source the AI can find.
Traditional keyword sets are usually quite small, but you’ve talked about a “Prompt Universe” that can include 30,000 questions. How can a human marketing team realistically manage or track a set of that size?
You don’t track all 30,000 individually; you use them to understand the “dialect” of your customers. The value of modeling a full Prompt Universe is that it reveals the sheer variety of how people ask questions—the SparkToro data showed us that 142 people could ask for the same thing and have almost zero semantic similarity in their phrasing. By seeing the whole universe, you can identify the clusters that actually matter to your business goals. You then choose a few hundred “well-chosen” prompts to track as a signal of your overall visibility. Because AI brand lists rarely repeat—there’s less than a 1-in-100 chance of getting the same list twice—trying to “rank #1” for ten prompts is just chasing noise. You want to see your brand appearing consistently across a broad spectrum of hundreds of queries; that is the only true signal of health in 2026.
What is your forecast for the future of the traditional B2B marketing funnel as AI continues to consolidate the research and decision-making process?
I believe the funnel is collapsing into a single, continuous dialogue. The traditional stages of awareness, consideration, and decision used to happen over weeks or months across dozens of different websites; now, they are happening in a single chat session that might last twenty minutes. A buyer can move from asking why they have high employee turnover to viewing a priced shortlist of software solutions without ever leaving the ChatGPT window. This means the “leakage” in our funnels will happen in places we can’t see unless we are actively monitoring the AI’s responses. My forecast is that “Brand Presence” will replace “Organic Traffic” as the primary KPI for B2B teams. If you aren’t the source the AI cites when it’s building that final shortlist, it won’t matter how good your SEO was in the past; you simply won’t be part of the final conversation with the sales rep. The vendors who map these invisible questions now will be the ones who define the category for the next decade.
