What marketers and brands need to understand about AI and buyer behavior

For most of the last decade, social media was a nice-to-have for B2B marketers — a channel for brand awareness, not a driver of purchase decisions. On the latest episode of After the Download, Kelsey Voss, Principal Marketing Analyst at eMarketer, argues that era is over. Buyers increasingly research, validate, and decide with input from peers, subject-matter experts, and AI systems that draw heavily on those same outside voices.

Voss has spent more than two decades in B2B marketing, leading teams inside SaaS and technology companies before moving to eMarketer, where she now leads a CMO series exploring how senior marketers are navigating AI, brand marketing, and go-to-market strategy. Her conversation with RealityMine® CEO Chris Havemann ranges across social media's growing role in the buying journey, the widening gap between what CMOs can see and what's actually shaping buyer decisions, and why AI adoption is running well ahead of AI maturity.

Here's what stood out.

Social media has moved deeper into the B2B buying journey

Social media used to be an awareness channel, Voss says — a way to introduce B2B buyers to a brand rather than to influence what they actually bought. That's changed. More of what buyers see now sits outside brand-owned channels entirely: peer networks, subject-matter experts, and industry voices they already trust to validate a purchase. AI systems draw from many of those same sources, which means marketers now need to work with the experts and creators shaping that conversation, instead of just publish more branded content.

Voss sees real parallels with consumer marketing but flags genuine differences too. B2B still lags behind B2C in working with influencers and creators, and longer, more complex sales cycles — often involving a full buying committee — mean more people across an organization are forming opinions on social media and through tools like ChatGPT, not just the named buyer.

CMOs are losing visibility into what's actually influencing buyers

Voss quoted a recent survey by Reddit and SurveyMonkey that found that 83% of US business decision-makers research on their own before ever speaking to sales. Many marketers say they can track activity through to closed revenue. But when asked how, only about half say they can do so with a complete attribution model.

She also mentioned a separate analysis by Meltwater, spanning six AI models and 16 B2B categories, which found that nearly half of the citations behind AI-generated answers came from social platforms, compared with less than 20% from company websites. Connecting that exposure back to pipeline, revenue, or even leads has become genuinely difficult, Voss says — and it only gets harder once a buying committee is involved, with each member picking up influence from a different source.

AI adoption is running ahead of AI maturity

Inside marketing organizations, Voss sees AI putting a renewed emphasis on data quality and governance, as fragmented systems and inconsistent customer data limit how effectively teams can personalize, segment, and analyze. Leadership pressure to show results with AI is real, but integrating it into existing martech stacks remains a major hurdle without change management and organizational alignment.

There's a lot of AI adoption, but not a lot of AI maturity yet.

The CMOs making the most progress, Voss says, tend to be specific about where they already trust AI to add value — and quick to pull back where it doesn't. One CMO halted an AI chatbot rollout after testing showed a 20% drop in satisfaction, redirecting resources back to fixing the underlying data issues instead.

Three trends B2B CMOs need to track over the next two to three years

Voss points to three shifts CMOs should be watching. The first is AI discovery: search rankings still matter enormously, but buyers are also using LLMs to discover and evaluate brands. That raises new questions for marketing teams — how a brand gets selected, how accurately it's being cited, and whether sales is representing it consistently once a prospect starts asking questions.

The second is the influence of outside voices — the creators, subject-matter experts, and peers buyers turn to for validation, whether or not a brand's own content ever appears in the sources an AI model cites to answer a question about it.

The third is internal: data governance and readiness. Voss argues that stronger foundations translate directly into a competitive edge.

The organizations with stronger data governance and stronger systems will be better positioned to integrate AI effectively.

GEO budgets are still a guessing game

Ask CMOs at even the largest companies how much they're planning to spend on generative engine optimization next year, Voss says, and the honest answer is often that they don't know. Her advice is directional rather than prescriptive.

Test and learn. Test and learn. Do not throw all your money at GEO.

Keep investing in paid social and search — still the two biggest budget lines — while using GEO to learn how buyers are actually engaging with AI models, rather than assuming a settled playbook already exists.

What this means for marketing teams, not just CMOs

Voss expects the bigger shift over the next few years to play out inside marketing teams themselves: who ends up owning GEO, how closely marketing coordinates with sales and product on consistent data, and how much data literacy becomes a baseline skill rather than a specialism. It's less about AI replacing marketers, she says, and more about new workflows — and new owners — for work that didn't exist a few years ago.

For B2B CMOs, the throughline of the conversation is less about any single AI tactic and more about visibility: understanding where buyers are actually forming their opinions, and building the data foundations to track it.

After the Download is hosted by RealityMine CEO Chris Havemann. New episodes explore how consumer behavior is changing in the digital economy — and what that means for the businesses competing for attention and share.

Listen to this episode on Apple Podcasts | Spotify | YouTube