10 Things Every Marketer Needs to Know About AI in 2026

AI is no longer a shiny experiment for marketing teams with time and budget to spare. It is now the operating layer of modern marketing. AI agents are building audiences, and conversational platforms are compressing the buyer journey into a single interaction. Measurement is moving from post-campaign reports to real-time optimization loops. Marketers who treat AI as a feature to evaluate are already behind the ones building on it as infrastructure.

The insights in this piece are drawn from more than 40 sessions at RampUp 2026, where leading practitioners shared how AI is reshaping the ways brands plan, measure, activate, and collaborate. Across every session, the same theme surfaced: the competitive edge in AI-powered marketing belongs to the brands that get the data, identity, and governance right.

These are the priorities that matter most right now.

Key takeaways

• AI performance depends on a strong foundation of connected, high-quality, and governed first-party data.
• Identity resolution and secure data collaboration make AI-driven targeting, activation, and measurement more accurate and actionable.
• AI agents will accelerate marketers’ work, but human oversight remains essential for strategy, governance, and decision-making.
• Marketers that move quickly while maintaining privacy, permissioning, and auditability will gain the strongest competitive advantage.

{{personalization-playbook}}

1. AI is only as good as your data foundation

Conversations around the potential of AI tend to land on the same truth: the model is only as good as what you feed it. Given incomplete, inconsistent, or poorly governed data, models provide incomplete, inconsistent results – faster than ever.

“If you just feed it noise, AI will amplify that noise,” said Thais Chang, Solution Engineering Chapter Lead at Infobip. Infobip tackled the noise problem directly. By partnering with LiveRamp, Infobip ties in-chat authentications and conversational signals to RampID, resolving fragmented campaign data into a unified, people-based view that brands can activate across channels.

Simon Whitcombe, Head of North America at Meta, reinforced the point about AI quality in a conversation with LiveRamp CEO Scott Howe: “These AI products are only as good as the data that is training them. And it’s the key ingredient.”

In his keynote, Howe made the competitive case. Public data is available to everyone, which means any advantage built on public data alone is temporary. Durable advantage comes from layering your own proprietary first-party data onto AI models. As Howe explained, when you stack unique data signals on top of what the public models already know, “you start a whole new series of S-curves and all of a sudden you jump up.”

Data collaboration makes this possible – connecting identity, insights, and signals into an AI-ready data layer that improves with every interaction. This starts with accurate identity resolution, scales through trusted partner connections, and deepens as your data grows richer.

2. Your proprietary data is your most important competitive moat

If the first point is about data quality, the second is about data ownership. In the age of AI, your data is not just an asset – it is your most valuable asset.

Howe put it bluntly: “In an AI world, data is power. Those that control the data will win, and those that do not will be captive to the model.” If you’re relying entirely on off-the-shelf models trained on public data, you’re competing with everyone else who is doing the same thing. 

Brand advantage comes from investing in permissioned data strategies and collaborating responsibly with partners to extract more value from the data you each have. Lauren Griewski, Head of Chase Media Solutions at JPMorganChase, argued that the signals brands choose to invest in matter as much as the models themselves: “For years in the industry, we’ve made decisions and optimized around proxies and clicks and modeled signals. [But] the most honest signal is the transaction, the verified transaction.”

Citi’s transformation tells the same story. By threading a common identifier through the full funnel, from targeting through measurement, the financial services brand built a connected view of the customer that became the foundation for smarter media decisions and measurably stronger acquisition performance. “Getting as full or complete a view of the customer as we can – we didn’t have that before,” said Kinjal Parikh, Head of Media Sciences at Citi.

LiveRamp helps you build that advantage. When you prioritize a durable identity strategy, your signals become more accurate, your AI models become more precise, and the ROI becomes measurable.

3. Agentic AI is an accelerator for humans, not a replacement

Will AI agents replace marketers? Will they run campaigns without oversight? These questions have swirled through marketing circles all year. The response from practitioners closest to the technology was: not yet, and not for a while.

“Today, and for the next couple years, AI is an accelerator for humans, not an independent operator,” said Aaron Andalman, Co-Founder and Chief Science Officer at Cognitiv. 

Ali Manning, COO at Chalice AI, added a caveat: “I think agents will become commoditized pretty quickly.” The differentiation, Manning argued, lies in the robust, AI-driven models that power the agents, not the agent wrapper itself.

Ravi Patel, CEO of SWYM, challenged the fear of agent-based displacement: “One thing that we need to know is they’re not going to replace marketers.” His practical advice for getting started? “Figure out in what areas you can use an agent to get you 70, 80% of the way there. You’re not going to have it just run wild, but it can help you become more efficient.”

The takeaway: AI agents are tools, and the marketers who deploy them effectively will outperform those who either avoid them or over-delegate to them. As Manning put it, “It’s not ‘AI’s going to take my job.’ It’s, ‘Oh wow, AI is elevating my job.’”

Through LiveRamp Agent Builders (LAB), marketers can license partner-built agents for audience building, measurement, and media optimization – and deploy them within the governed data collaboration environment they already trust.

4. Without identity as the backbone, AI is a fantasy

AI needs to know who it is talking about. If you lack a consistent, accurate identity framework connecting your data across channels, partners, and platforms, all your AI-driven insights are built on sand.

Erin Mitten, VP of Product Strategy at Publicis, was direct: “I think it’s a fantasy if you don’t have deterministic identity driving it.” Without that deterministic foundation, she explained, what you get is “probabilities on top of probabilities that have compounded into something you can’t even begin to track back.”

Domenic Venuto, EVP and Chief Product and Data Officer at Horizon Media, saw opportunity for brands: “The promise of AI is that it cuts through all of that, brings it together, and actually finally connects all the dots.” He was referring to the fragmentation that has long plagued cross-channel measurement, audience planning, and personalization. AI can solve it, but only if there is a consistent identity layer underneath.

RampID, LiveRamp’s durable, interoperable identifier, is the connective tissue that makes AI-powered marketing addressable and measurable. It links authenticated identifiers across the digital and martech ecosystem, so your data collaboration, activation, and measurement all operate from the same customer view.

5. Governance and trust are competitive differentiators, not speed bumps

One of the more surprising themes at RampUp 2026 was how often governance came up – not as a compliance burden, but as an advantage.

John Hoctor, CEO of Newton Research, issued a pointed warning on agentic measurement: “The digital publishers have learned a lesson about these big AI companies and how they treat intellectual property, scraping all that information and now that’s built into their models. I’d be very, very careful with your corporate data.” Newton advises deploying agents as containers inside the client’s own cloud environment. “None of the data ever leaves your environment,” Hoctor explained.

Manik Khanna, CEO of SemantIQ Health, reinforced this from the healthcare perspective: “In a regulated industry like healthcare, auditability is extremely important.” SemantIQ deploys its agents within the LiveRamp Clean Room, ensuring full transparency into data lineage with every query.

The brands that build governance into their AI workflows from day one – rather than bolting it on after something goes wrong – will earn the trust of consumers, partners, and regulators. That trust builds over time: partners are more willing to share high-value data when they know the governance is real, and consumers are more willing to share first-party information when they trust how it will be used.

LiveRamp is built to enable end-to-end control: advanced permissioning, consumer protections, and data ethics standards are available to brands and platforms at every layer, not added as an afterthought.

6. AI compresses the buyer journey, so every touchpoint is now a conversion moment

The traditional marketing funnel – awareness at the top, conversion at the bottom, a neat handoff in between – is collapsing. AI-driven experiences are compressing the journey from discovery to purchase into a single interaction.

“Conversational AI platforms do compress that journey from awareness to consideration to a decision to even purchase, all in one place,” said Debra Aho Williamson, Chief Analyst at Sonata Insights. She cited McKinsey research showing that 73% of US consumers want to use AI to learn about products and services, 60% are already using AI tools to explain features and technical specs, and 61% are using AI to compare specific products.

Mark Balabanian, Chief Business Officer at Dappier, shared his experience: ads within conversational AI experiences get “a response rate that is like three to five times the response rate of a comparable ad on a webpage.” Users are not passively scrolling. They are engaged in a two-way exchange, paying attention, and in an active mode of participation.

For Dave Osborne, GM of Agentic Solutions at Microsoft Advertising, the shift represents a structural opportunity: “What I would like to see happen is I would like to see discovery beat dominance.” The promise of AI, Osborne argued, is democratization – giving smaller and local businesses the chance to be surfaced based on relevance to a consumer’s needs rather than sheer media weight.

Every interaction with a consumer, whether it happens in a chatbot, a search result, or a conversational AI surface, needs to be optimized for conversion, not just awareness. The teams that still plan media around a linear funnel are optimizing for a journey that fewer and fewer consumers actually take.

7. New AI surfaces are the next addressable media frontier

ChatGPT, Perplexity, Google AI Overviews, and a growing number of conversational AI platforms are becoming primary discovery channels, and they need to be treated as media.

“Frankly, every publisher needs an AI mode,” said Mark Balabanian of Dappier, explaining that “AI is a new medium. It’s a new opportunity.” The shift is already well underway. LiveRamp recently announced a measurement partnership with OpenAI, enabling brands to track conversions and optimize media performance across ChatGPT’s advertising surfaces – the same capabilities marketers expect from established channels.

For marketers, the opportunity is twofold. First, these AI surfaces represent incremental reach – consumers who are bypassing traditional search entirely. Second, they are measurable. With RampID connecting AI-influenced interactions back to a durable identifier, you can understand which conversations led to which outcomes and optimize accordingly.

The brands that treat conversational AI as a media channel today will be the ones ready with performance data and optimization playbooks when it scales.

8. Measurement is shifting from post-campaign reports to real-time optimization

For years, measurement has been a backward-looking exercise – useful for learning, but too late to change anything. AI is giving marketers the chance to turn measurement into a live input. 

John Hoctor of Newton Research shared his firsthand experience: “For a long time I was in the report card business.” Now, with agentic AI operating on top of LiveRamp’s cross-media intelligence environment, Hoctor sees a different model emerging. “I like to joke that with measurement, it’s like trying to drive your car through the rear view mirror,” he said. “You really want to figure out…where to go in the future, not just look behind at what happened.”

Newton’s agents don’t just surface insights. They can take action, placing media buys programmatically through platform APIs. “You’re breaking out of measurement as a report card,” Hoctor explained. “You're really turning measurement into an input into that closed loop cycle.”

PJ Gasparini, SVP of Insights and Measurement at NBCUniversal, provided the media owner’s perspective. NBCU’s new Performance Insights Hub (PIH), co-developed with LiveRamp, automates the flow of campaign data into a clean room on a weekly cadence for measurement and optimization. “We’re shifting from post-campaign insights to always-on in-flight,” Gasparini said. “And that’s a really core component of PIH, which is giving consistent feedback and enabling that more near real time decisioning.”

LiveRamp’s measurement and analytics capabilities give you not only the infrastructure for dynamic, AI-driven optimization, but also the identity layer to make every insight actionable.

9. Data collaboration is the multiplier

No brand has all the data it needs. The ones that recognize this and invest in secure, governed data collaboration are unlocking exponential value from signals they could never access alone.

Scott Howe offered the example of Dick’s Sporting Goods. With 45 million athletes, more than 600 attributes, and over two billion touchpoints across the customer journey, Dick’s collaborated with Under Armour and Roku on a campaign that reached nearly three million athletes and delivered a $4.18 return on ad spend. That result was powered by data collaboration, measured through clean rooms, and protected at every step.

Haleon, CVS, and Reddit partnered to combine contextual signals from Reddit’s health communities with CVS’s deterministic purchase data. The result, measured through the LiveRamp Clean Room (with appropriate privacy guardrails that met each customer’s guidelines), was 600% higher unit sales growth for the category during the closed beta versus the same period the year prior. The signals that drove that outcome did not exist inside any one company’s data set. They emerged from collaboration.

These examples illustrate a core principle of data collaboration: new signals become exponentially more valuable when connected to a trusted network of high-quality partners. One partner’s contextual data, combined with another’s transactional data, combined with a third’s behavioral signals, produces insights none of them could have generated independently.

LiveRamp’s data collaboration network spans more than 800 customers and 500-plus top-quality partner integrations. It’s the infrastructure behind these partnerships – with the governance and permissioning to ensure every collaboration is secure, auditable, and aligned with consumer expectations.

10. Your biggest opportunity lies in moving fast and governing well

There is a persistent myth in marketing that speed and governance are in tension – that you can move fast, or you can be responsible, but not both.

“Buckle your seatbelt, but go as fast as you can, otherwise you become irrelevant,” said Domenic Venuto of Horizon Media. He was not suggesting recklessness. He was arguing that the infrastructure for responsible speed already exists, and that brands failing to adopt it risk falling behind.

John Hoctor underscored the urgency from a resourcing perspective: “We’re all going to be asked to do more with less. We’re all going to be collecting more data. Trying to get all that value is next to impossible if you’re not leveraging AI and agents.”

Speed without governance creates brand safety risk, data misuse, and regulatory exposure. Governance without speed leads to irrelevance. Successful marketers in 2026 are doing both – deploying AI-powered workflows that accelerate execution while maintaining the data ethics, permissioning, and auditability standards that protect their brand and their consumers.

LiveRamp can support you on this mission. Our platform delivers AI-driven marketing performance with governance controls embedded throughout the workflow. This is how you achieve both trust and scale – and it’s the defining competitive advantage of the AI era.

{{personalization-playbook}}

Data collaboration for a strong AI foundation

AI is the new operating system for marketing – and the brands that treat it as such are already pulling ahead. The 10 takeaways above are not predictions. They are the lived experiences of the most sophisticated marketers in the industry who are deploying these capabilities at scale today.

The common thread across all 10? Data collaboration. Every AI-powered outcome – from sharper measurement to smarter agents to more relevant consumer experiences – depends on connected, governed, high-quality data flowing across a trusted network of partners.

LiveRamp is that network. We help you connect every signal, every agent, every surface, and every channel – so your AI-powered marketing delivers performance you can measure and trust you can build on.

Ready to build your AI foundation? Here’s where to start.