Hi — Asad here. Before we get to today’s editorial, I need to tell you about the conversation we just had with Amanda Kahlow, CEO of 1mind.

1mind sells a “digital superhuman” — think: AI salesperson. Now, I’m a purist who believes that salespeople, with all their imperfections, will always beat a machine. But I hold that opinion loosely, and Amanda is the kind of person who makes you hold it even more loosely. She told us exactly what’s working today, where it’s going — and was crystal clear that she doesn’t see sales jobs existing in the future. Uff.

Less than 3% of AI chat sessions are transactional. 

Most marketers look at that stat and see a reason to ignore the channel. They’ll choose to continue to invest in old, broken tactics: paid social = saturated; SEO returns = flattening; programmatic = a tradeoff between reach and user experience.

But I believe that’s a mistake. Because the other 97% of those sessions is where buyers are doing the research, the comparisons, and the “help me figure out what to buy” conversations that used to happen on Google and review sites, and in analyst reports.

Your customers have quietly migrated to this new interface — which doesn’t serve banner ads, doesn’t reward keyword stuffing, and doesn’t hand you a click-through rate. And, if you’re not testing into this now, you’re not showing up where those key decisions are actually being made.

If AI is where people search and, perhaps more importantly, where people think, then we’ve all got to stop trying to capture clicks, and instead, use what we know about our customers and their contexts to shape decisions where they’re actually being made: inside AI conversations.

The sales funnel that most SaaS marketers grew up with still exists. You’re still trying to build awareness, gain consideration, drive conversions. Except now, the moments inside it have scattered.

So, for example, a VP of Engineering who’s evaluating infrastructure tooling might start in ChatGPT, pivot to a Slack thread, skim a comparison blog, and return to AI to synthesize all they’ve learned before ever filling out a form. 

In today’s ecosystem, AI interfaces are at the center of everything, acting as researcher, recommender, and filter. In many cases, entire portions of the decision-making process are compressed into a single AI session.

The implication is both simple and uncomfortable: Those moments of intent that we’re all going after are more elusive, because they’re now living inside AI responses. So we’ve got to work harder to capture them, to drive the conversation in our favor.

Above all else, the best ad is the one that helps someone decide.

Traditional ads appear around content. They’re interruptive, audience-based, click-driven.

In contrast, AI-native ads appear inside responses. They’re assistive, context-based, and decision-driven. They show up because the query made them relevant, not because a user matched a predefined audience segment.

We’ve all been conditioned to traditional marketing channels and the kinds of ads that run on them. We need to think about AI differently — not as another channel but as a new surface area that runs on trust. 

In AI, usefulness builds trust, and anything that feels invasive erodes it just as quickly. Take these two Super Bowl campaigns: Anthropic’s “What Do You Think of My Business Idea?” ad positioned Claude as a collaborator, drawing a clear line between helpful AI and experiences that feel too commercially driven. Compare that to Ring’s AI-driven “Search Party” dog-finding program, which drew backlash for crossing privacy boundaries. 

Marketers who treat AI carelessly will only accelerate the kind of backlash that kills emerging channels, while those who get it right will meet buyers earlier, and more meaningfully, than ever before.

Advertising has lived and breathed through audience personas, but no more. That’s because traditional personas are simply too blunt for AI environments.

Take this example: “VP Marketing at a 500-person SaaS company.” Sure, this tells you who the customer is. In the AI world, that’s hardly going to move the needle.

Instead, we need to shift our targeting logic from identity to intent, integrating queries, topic clusters, and decision-stage signals into the language of your brief — “Comparing CRM options for a mid-market sales team.” 

This way, you get closer to what the customer needs right now, vs. simply who they are.

AI compresses the funnel, but more importantly, it pulls decision-making earlier in the process. What used to be top-of-funnel research is now where buyers are actively forming opinions, comparing options, and narrowing their shortlist.

Rather than running full-funnel campaigns end to end, you’ve got the opportunity to show up in those early decision moments: comparison queries, trade-off questions, and “what should I choose?” prompts.

That’s why a cybersecurity vendor gets more leverage appearing inside “best endpoint protection for a remote-first team” than from a generic, top-of-funnel campaign — because in AI, that’s the decision moment.

Remember: The bar you’re aiming for is usefulness, not persuasion.

The best-performing AI ads go beyond promoting something, to provide something useful: a specific insight, a clear use case, or a meaningful point of differentiation.

For instance, a project management platform that responds to “how do mid-market ops teams reduce meeting overhead?” with a concrete workflow — and then surfaces its product as the tool that enables it — is doing this right.

Here’s a simple formula to keep top-of-mind: context → insight → suggestion.

AI enables 1:1 messaging at scale — the holy grail that SaaS marketers have been seeking for decades. Creative, value props, and calls to action can all be generated dynamically based on the query, context, and user need.

This means that a sales intelligence platform worth its OpEx should be able to match this level of personalization by, for example, showing a different message for “prospecting enterprise accounts” than for “building pipeline in a new vertical.”

Most advertisers aren’t doing this yet. The ones who figure it out early will have a meaningful advantage.

We’ve all been trained to get the clicks. But clicks go underrepresented in AI environments: A user may read a recommendation and not click, but come back and search your brand 10 minutes later. Though that’s still a win, the click model won’t capture it.

AI advertising is closer to sales enablement than media buying — and should be measured like it. We’d all be better off with a performance measurement  model that accounts for:

This is non-negotiable: Trust isn’t a feature of AI advertising. It’s the foundation.

AI ads must be relevant, clearly labeled, and non-intrusive. If ads degrade the experience, the channel degrades with it.

There are two ways to show up in AI: 

SEO for AI is a long-term play, best used for situations where authority and coverage matter over time. It’s how you become part of the default answer when someone asks fundamental questions like “what’s the best CRM for mid-market teams?” or “how should I structure a demand gen team?”

In contrast, SEM for AI is immediate, and best when buyers are actively narrowing options. It lets you show up inside high-intent moments — comparison queries, trade-offs, and decisions like “HubSpot vs Salesforce” or “how to evaluate endpoint security vendors” 

Each has their own place in a campaign, and the most effective teams use both: SEO to earn presence, and SEM to capture decision moments.

Doing AI-native advertising well doesn’t require a wholesale budget shift.

My suggestion: Begin by reallocating 5–10% of an existing budget — ideally from a channel that’s underperforming — into AI-native testing.

You need to think of AI as a new decision layer, and thus, make your campaign meaningful. Otherwise, you’re unlikely to generate enough signal — which could lead you to write the channel off too early, underinvest, and miss the window to build an advantage.

These steps can help you to assess your advertising investment:

AI is having as profound an effect on marketing as every other GTM motion: Advertising is no longer just about generating demand. (Let’s be honest: It hasn’t been for a while now, but…) Moving forward, it’s more about shaping decisions where they’re actually being made — reaching people, and adding value on behalf of your brand while they’re in the best position to buy what you’re selling.

Today, all of that is happening inside AI.

The buyers have already moved. The only question is whether marketing will adapt and catch up.

Kane Russell is a GTM leader with 20 years of experience scaling B2B growth through major platform shifts, from SEO and mobile to SaaS and AI. He has led GTM for companies backed by A16Z, Sequoia, Forerunner, and Theory Ventures. Currently CRO at Koah, he focuses on how generative AI reshapes demand generation and buyer intent.

Become a Topline insider by joining our Slack channel.

We want to hear your feedback! Let us know your thoughts.