SAP SE announced it has completed the acquisition of Prior Labs, the pioneer of Tabular Foundation Models (TFMs), accelerating SAP’s success in TFMs, which began with SAP-RPT-1, and bringing one of the world’s leading TFM research teams into the SAP family.

Prior Labs will continue to operate as an independent entity, with SAP committing to invest more than €1 billion over the next four years to scale it into a globally leading frontier AI lab for the structured data that runs the world’s businesses. Terms of the deal were not disclosed.

Large language models (LLMs) struggle to make accurate predictions on structured business data because they have only a rudimentary understanding of tables, numbers, and statistics. Unlike LLMs, TFMs are purpose-built for this type of data and can accurately predict business outcomes from tabular data, such as payment delays, supplier risks, upsell opportunities, customer churn risk, and more.

Also Read: AI Should Remove Friction, Not Human Connection

“Early on, SAP recognized that the greatest untapped opportunity in enterprise AI wasn’t large language models; it was AI built for the structured data that runs the world’s businesses,” SAP CTO Philipp Herzig said. “We built SAP-RPT-1 to prove that conviction for enterprise data. Prior Labs has built a leading TFM on public benchmarks and built one of the leading research teams in this category. Combining their frontier model work with enterprise data and customer reach is how we intend to lead this category globally.”

“Over the last 18 months, Prior Labs has built an incredible team, increasing the velocity in tabular foundation models,” Prior Labs CEO Frank Hutter said. “Joining the SAP family gives us the resources, data environment, and customer reach to take this category to its full potential.”

With Prior Labs now part of SAP, the two companies have the opportunity to establish an industry-leading AI research lab and shape a new category in TFMs. The lab will operate as an independent unit to ensure research velocity, while SAP provides long-term investment and a direct path to productization across the SAP portfolio, including SAP AI Core and SAP Business Data Cloud, as well as the agentic layer with Joule.

With over 3 million downloads, Prior Labs’ TabPFN is a widely adopted open-source tool for tabular AI, supporting a dynamic developer ecosystem. SAP is fully committed to continuing to support this open-source strategy. Prior Labs cofounders Frank Hutter, Noah Hollmann, and Sauraj Gambhir lead a team of world-class AI researchers and practitioners. The company works with leading scientists in the field, including Yann LeCun, ACM A.M. Turing Award winner and executive chairman at Advanced Machine Intelligence, and Bernhard Schoelkopf, director of the Max Planck Institute for Intelligent Systems and ELLIS president, both of whom will serve on Prior Labs’ scientific advisory board as it scales into a globally leading frontier AI lab.

Also Read: HubSpot’s Aja Frost on Marketing in the Age of AI Search

Accelerating Innovation

Prior Labs’ TabPFN-2.6 is the top-performing model on TabArena, the leading benchmark for TFMs. TabPFN-2.6 matches the accuracy of a four-hour automated machine learning pipeline — instantly, in a single model, at a fraction of the complexity.

With a conversational interface layered on top, business users can ask questions in natural language, generate or select datasets, and run “what-if” scenarios without needing to be data science or machine learning experts. With Prior Labs’ models, SAP will provide in-context learning, allowing users to supply data records and receive instant, reliable predictions without any model training. A single TFM can adapt to any business use case on the fly, delivering faster time-to-value while ensuring GDPR compliance.

With Prior Labs, SAP will deliver TFMs with superior predictive capability that understand tables natively, learning statistical reasoning directly from data, and will power agentic AI systems capable of understanding high-level goals, combining tables, language, and images to reason, integrate domain knowledge, infer causality, and adapt dynamically.

Now that the deal has closed, SAP and Prior Labs plan to turn top AI research into enterprise-ready innovation, enabling customers to get even more value from their tabular business data. True intelligence requires moving beyond correlation to understand causation. Answering “What will happen?” is useful, but answering “Why will it happen?” is transformative.

The post SAP Completes Acquisition of Prior Labs to Lead AI appeared first on MartechView.

Three-quarters (75%) of B2B CEOs and senior business leaders don’t believe marketing drives business growth, even though they acknowledge it plays an important role in their organizations.

That’s according to The CEO Blind Spot, a new report launched today by Propolis, which surveyed 150 UK CEOs and business leaders at B2B organizations to explore how they perceive marketing’s contribution to commercial success.

The report identifies what it describes as the “CEO blind spot”: a tendency for organizations to attribute growth to the point at which revenue is recognized, while overlooking the months of demand creation and brand building that often make that sale possible.

The findings reveal that 84% view marketing as a support function rather than a driver of commercial growth. That perception is reflected in how business leaders compare marketing with sales: 77% believe sales is a bigger driver of growth than marketing, while 67% believe marketing is less accountable for business results than sales.

The research also suggests these perceptions are beginning to influence investment decisions. More than a third (35%) of business leaders believe marketers are being held back at the board level because investment is increasingly being directed toward innovation and AI.

Richard O’Connor, CEO of Propolis, said: “Too many CEOs say they value B2B marketing, but our research suggests they still don’t value it as a commercial growth function. If you believe marketing matters but doesn’t drive growth, it’s difficult to argue you’re recognizing its full contribution to the business.

Also Read: AI Should Remove Friction, Not Human Connection

“The challenge is that much of marketing’s commercial contribution happens long before revenue appears on a dashboard, making it far less visible than that of functions operating closer to the point of sale. As CEOs face growing pressure to deliver short-term results while increasing investment in AI, there is a real risk that a critical engine of sustainable growth becomes an easy target for budget cuts unless this blind spot is addressed.”

The full CEO Blind Spot report is available for download here, offering further insight into why marketing continues to struggle for recognition as a commercial growth driver and what that means for B2B organizations.

The post 75% of CEOs Don’t Think Marketing Drives Growth appeared first on MartechView.

Pedro Andrade has a simple way of explaining why he took the VP of AI role at Talkdesk. The contact center, he says, has been broken for decades — and everyone in the industry knows it. Agents are navigating ten legacy systems simultaneously. Customers are on hold while repeating information they already submitted online. A department that exists, in most organizations, to absorb complaints rather than prevent them.

What changed his mind about the timing wasn’t a single breakthrough. It was the convergence of two things: a decade of failed deployments that taught the industry what actually matters, and models that have finally caught up to the promises vendors were making years too early.

Andrade splits his time between engineering teams debating RAG pipelines and latency, enterprise CIOs who are simultaneously excited and terrified about AI, and a forward-looking 30% of his calendar dedicated to asking which foundation models are worth betting on twelve months from now.

In this conversation, he talks about what separates real AI from the chatbot era’s false promises, why Klarna’s much-publicized reversal was actually a story about process design rather than technology, and why the most important question in AI-powered customer experience has nothing to do with the model.

Excerpts from the interview; 

What drew you to Talkdesk, and why is AI in customer experience such an interesting problem to solve today?

CX has been broken for decades, and everyone knows it. Contact centers are perceived as cost centers. Agents burned out navigating ten legacy systems. Customers are waiting and repeating themselves.

I took this role because Talkdesk sits exactly where this gets fixed. I’m lucky to live in a time when we have “enough technology” genuinely capable of resolving this problem. Not just deflecting calls to save money, but turning the contact center from a complaint department into something proactive and revenue-generating. Making both the agent’s and the customer’s lives meaningfully easier is, honestly, one of the most interesting problems in enterprise tech right now.

Also Read: The Agency-Led E-commerce Model Is Changing

What does your role as VP of AI actually involve? Where do you spend most of your time, and what are you building?

The title sounds futuristic, but the day-to-day is pretty pragmatic.

About 30% of my time is with engineering and data science, translating what new models can do into features that actually work in production to solve real pains (there are always pains somewhere, even in advanced tech). That means architecture reviews for our AI Agents, RAG pipelines that don’t hallucinate, and a deep commitment to latency.

Another 40% is with customers. Enterprise CIOs are excited about AI and terrified of it at the same time. I help them build realistic roadmaps that align with their data constraints and brand. It’s as much change management as it is technology.

The remaining 30% is looking ahead. The landscape shifts fast enough that if I’m not thinking 12 months out, we’re already behind. Which foundation models are worth betting on? How do multimodal, new hybrid models, etc, change the customer interaction model? That’s where our R&D decisions get made.

Contact centers have used AI for years—from chatbots to IVR. What’s fundamentally different about what Talkdesk is building today?

The honest answer is that we finally have technology that’s up to the promise. For years, “AI in the contact center” meant rules-based bots that frustrated customers and sentiment scores nobody acted on. The gap between what vendors claimed and what actually worked was enormous.

That gap hasn’t closed because we got smarter overnight. It closed because a decade of failed deployments taught us what actually matters, and the underlying models finally caught up. You can’t skip that learning curve.

What’s also changed is the scope of what’s possible. Eight years ago, during the chatbot boom, nobody seriously imagined we’d have this level of language understanding, let alone systems that can process images and video and reason about what they mean. That expands the use cases dramatically, and honestly, it creates a whole new set of problems to solve. Which is the part I find most interesting!

When should an AI agent hand a conversation to a human—and who decides where that line is drawn?

The client always owns the decision. Talkdesk provides the engine, but the enterprise dictates the business logic.

Also Read: AI Ads Will Win Only If They Earn Consumer Trust

Every CX platform is making big AI claims. Where does Talkdesk genuinely stand apart in real-world enterprise deployments?

A few things differentiate us, and I’ll be direct about what I think actually matters versus what’s table stakes.

The data advantage is real. Years of interaction data, combined with deep integrations into enterprise systems of record, give our models context that a generic LLM simply doesn’t have. The 360-degree customer view isn’t a marketing line; it’s what makes automation actually work in production and provide truly personalized content. 

We’re also not selling a bot. We cover the full contact center surface: self-service, agent assist, routing, workforce management, outbound, analytics, and back-office automation. That end-to-end scope matters because CX problems don’t live in one place.

Industry specialization compounds that. Prebuilt vertical agents and industry-specific retrieval models mean we’re not starting from zero with every customer. That’s what lets us go from pilot to production in days or weeks rather than quarters. 

Two things I’m particularly proud of. 

• First, AI Gateway lets our AI run on top of third-party or on-prem environments, not just Talkdesk. Many enterprises are nowhere near full cloud migration, and meeting them where they are is a competitive reality. 
• Second, differentiation also takes the form of go-to-market motions: we deliberately shifted toward forward-deployed engineers co-building with customers. That changes the quality of what gets shipped, and puts Talkdesk as co-owner of the solution we deliver to customers. And they love it.

And underneath all of it: enterprise-grade trust and safety. Guardrails, governance, human-in-the-loop controls. Our enterprise banking customer taught us early that this isn’t optional for the deals that matter.

Klarna replaced hundreds of support agents with AI, then began hiring humans again. What did the industry misunderstand from that experiment?

Klarna is a great case study, and I say that without judgment. They moved fast, made a bold bet, and learned from it publicly. That takes courage.

What the industry got wrong was treating this as a technology swap rather than a transformation. You can’t just replace humans with AI and declare victory. A lot of what makes customer experience work has nothing to do with the model: it’s data readiness, process design, change management, and organizational culture. Those differ from company to company, and no vendor can package them for you.

The other mistake was skipping the learning cycles. This technology is genuinely new. The right motion is to start small, instrument everything, and treat early deployments as experiments with real feedback loops, not finished products. Companies that went all-in before building that foundation set themselves up for exactly the kind of reversal Klarna experienced.

The ones getting it right are moving incrementally, measuring outcomes honestly, and accepting that some human judgment isn’t a failure of automation. It’s a feature.

Also Read: Your AI Chatbot Knows More Than Your Marketing Team

Some argue that companies are using AI to avoid customers rather than serve them. As someone building these systems, how do you respond?

At Talkdesk, we design AI to remove friction, not to remove connection. If a customer just wants to know where their package is, making them wait 15 minutes on hold for a human is disrespectful of their time; AI solving it in 10 seconds is excellent customer service. But if they need help navigating a missed mortgage payment, forcing them to talk to a bot is cruel. The technology isn’t the problem—the deployment strategy is.

The post AI Should Remove Friction, Not Human Connection appeared first on MartechView.

Credit card rewards were built for a world where the issuer set the terms, the consumer spent, and the value arrived later.

But as payments sped up, that legacy model stayed the same. Smart basket technology could help close the gap. Findings in “The Smart Checkout Opportunity: Why Issuers Are Ready for a Better Rewards System,” a PYMNTS Intelligence report produced in collaboration with FIS, reveal that nine in 10 issuers surveyed are highly interested in adopting smart basket systems. More than eight in 10 say they would participate if such a system became commercially available.

And across all cardholder behaviors studied, issuers rated smart basket tools far more effective than current rewards, with net effectiveness scores ranging from 43 to 59 percentage points.

But that enthusiasm should not be mistaken for simplicity. The report’s most revealing finding is not that issuers want a smart basket. It is that they want it on tightly governed terms. Issuers view smart baskets as a way to reclaim control over loyalty economics. Three in four expect such systems to give them more control, especially over targeting, merchant participation, and offer timing.

The word “control” matters. As rewards shift from broad issuer-funded programs to incentives funded by brands, merchants, and other ecosystem participants, issuers are not simply adding a new feature. They are deciding how much of the cardholder relationship, data flow, and incentive economics they are willing to expose to outside partners.

Traditional Reward Offers Have a Relevance Problem

Traditional rewards still work reasonably well as blunt instruments. They can encourage card use, lift spend, and support broad engagement. But they perform less well where issuers increasingly need precision.

Fewer than half of issuers say their current incentives drive loyalty or retention. Only 47 percent say rewards cause cardholders to switch to their cards, and just 40 percent say they shift the timing of spend.

Smart basket technology attacks that weakness by making rewards immediate, contextual, and transaction-specific. Instead of asking consumers to activate an offer in advance or wait for a statement credit after purchase, the system can identify eligible savings in real time and apply them at checkout. The value becomes visible when the consumer is deciding how to pay.

That matters because rewards are no longer competing only against other rewards programs. They are competing against friction, relevance, and timing.

Issuers see Smart Basket as a way to regain control over loyalty economics, particularly across the data that determines eligibility, the rules that govern timing, the funding flows that settle the offer, and the liability model that determines who pays when the system fails.

The report highlighted the following breakdown:

• Rules: Smart basket only scales if issuers can govern the transaction layer — who qualifies for an offer, when it appears, who funds it, how disputes are handled, and where liability sits when incentives misfire.
• Data: The value shifts from generic rewards data to permissioned, transaction-level intelligence that can identify basket contents, cardholder behavior, merchant context, and incremental lift in real time.
• Economics: Rewards move from broad issuer-funded subsidies to a more dynamic incentive marketplace, where brands, merchants, and issuers co-fund targeted offers tied to measurable spend, retention, and conversion outcomes.

The card rewards war is therefore entering a more technical phase. Richer offers will still matter. So will consumer experience. But instead of rewarding entire categories with static rates, issuers could participate in a system in which brands and merchants fund targeted behavior at the transaction level. This turns the path forward into one that goes straight through the incentive layer inside the transaction.

Smart basket may arrive as a better way to save at checkout. Its deeper impact is that it turns loyalty into infrastructure.

The post Issuers Want Smart Basket Rewards, on Their Terms appeared first on MartechView.

AGNT LAB has launched a social media AI agent aimed at entrepreneurs and small businesses nationwide. The agent’s features and enhancements continue to be updated across freemium, pro and max user categories. From simple schedulers to full multichannel agents, the three-tiered options help with everyday social media management tasks.

“At AGNT LAB, we are leveraging the speed and reasoning of AI, allowing small businesses, as well as startups, an easy way to get their social media going and keep in the loop, during even the busiest of days,” said Jamahal Winston, founder of AGNT LAB.

AGNT LAB automates a brand’s social media using context-aware AI agents that create, schedule, engage and convert. The company is AI-powered to automate organizational processes seamlessly.

The freemium version allows an entrepreneur or small business owner to schedule posts in one channel and monitor comments and direct messages with a history of availability. The agent works across Facebook, Instagram, X and LinkedIn, with TikTok access available by the end of August. Beta users who have been onboarded for more than 90 days can upgrade to a discounted paid plan with additional capabilities.

Also Read: Your AI Chatbot Knows More Than Your Marketing Team

The pro version includes the ability to add more agents and channels, and to manage lead tracking with key performance indicators and analytics. At the max version, the highest tier of service, the agents can respond to lead requests by generating outreach content after a lead is received, performing sentiment analysis of the lead and showing the number of views on posts.

“I’ve used the agent to support my social media posts, keeping me on track, and AGNT LAB has contributed to my productivity on a weekly basis. I can’t wait to see what they will do next,” said a beta user.

“Because we are AI-powered, unlike other social media platforms on the market today, we assist in the actual content generation phase, keeping it easy and managing content. You set the agents’ autonomy, and they are supervised. For every action, you are in complete control, and the system will ask permission to move to the next step, as well as learn your brand voice,” Winston added.

Also Read: Are We Using AI to Help Customers or Avoid Them?

Like other social media managers, AGNT LAB will auto-connect to a user’s channels after the first session and review past content to establish brand voice, he said. The agent or agents will also have access to a user’s image and graphic library.

The post AGNT LAB Launches AI Social Media Agent for Small Business appeared first on MartechView.

Most attractions and destinations still treat AI chat as a support tool. Something that answers hours and ticketing questions so the phone line rings less. That view misses what is actually happening in these conversations. Every question a visitor asks an AI agent is a signal about what they want, when they want it, and what would get them to spend more. Organizations that treat that signal as a customer service log are sitting on data that marketing teams would pay for.

A good example of this comes from Franklin Park Conservatory and Botanical Gardens in Columbus, one of the founding members of the Agentic City program launched by Satisfi Labs this year with Experience Columbus. Over a three-month period, Franklin Park’s AI agent fielded thousands of guest conversations, and the patterns inside those conversations tell a much bigger story than “we answered some questions.”

Start with timing. A large share of the conversations occurred at 10 pm, 11 pm, and later hours, when no staff member is on site to answer the phone. People were asking about hours, ticket availability, membership eligibility, and event registration well after the gates closed. For an attraction, that is a staffing problem solved without adding staff. For a marketer, it is something else entirely. It is a record of intent that would have evaporated by morning.

Also Read: HubSpot’s Aja Frost on Marketing in the Age of AI Search

That intent shows up most clearly around events. When Franklin Park hosted a Gabby’s Dollhouse Meet ‘n’ Greet, the agent absorbed a massive spike in volume, easily a third of all conversations during that window. Guests asked how to register, why a second ticketing step wasn’t appearing, and whether the event was sold out. The agent guided families through a process at scale, without additional hires. But it also surfaced something the events team hadn’t had visibility into before. Teams could see exactly where the ticketing flow was breaking down for real guests, in real time, and in their own words. What began as a customer service interaction ultimately became a source of insight for both the events and marketing teams.

The same pattern revealed something traditional analytics tools rarely catch. Guests asked repeatedly about Museums for All and social services-based discount programs, including questions about specific managed care plans and out-of-state eligibility. These are not casual browsers. They are people actively trying to visit the Conservatory, and the questions show how conversational AI can help visitors navigate access programs and find information that might otherwise be difficult to obtain. That is an audience insight with implications for outreach, programming, and community partnerships that extend well beyond admissions.

Then there is the revenue conversation hiding inside the customer service conversation. Guests asked about weddings, birthday parties, graduation parties, and barn rentals, often with specific dates, venues, and budgets in mind, and frequently at night when the events team was unreachable. Others asked whether the special exhibit changes seasonally, when a specific show is running, and how much the butterfly larvae in the gift shop cost. None of that would show up in a sales report. It shows up only in the conversation itself, and only if someone is positioned to capture it.

Also Read: The Agency-Led E-commerce Model Is Changing

This is the shift marketers need to make. Unlike traditional FAQ tools, AI-powered conversations do more than reduce support volume. They’re giving marketers direct visibility into audience intent. A guest asking about wedding venues at 10 pm is a lead. A guest’s question about whether an exhibit rotates seasonally is a programming signal. A guest asking about a discount program is an equity and outreach signal. A spike in confused questions about a single event is a UX problem marketing can fix before the next campaign.

Every one of these signals already exists inside the conversations attractions are having with visitors right now. Marketers have spent years trying to infer customer intent from clicks, page views, and conversion data. Increasingly, customers are telling us exactly what they want. The organizations that learn to listen will have an advantage that no dashboard can provide.

The post Your AI Chatbot Knows More Than Your Marketing Team appeared first on MartechView.

Two hundred and fifty years after the signing of the Declaration of Independence, Google has released a new commercial asking a very 2026 question: what if the Founding Fathers had access to Google Workspace?

With the tagline “Group project, but make it 1776,” the ad depicts a largely unseen Thomas Jefferson mid-draft, receiving a nagging text from Ben Franklin that sets off a very Google-centric collaboration process. Edits are suggested in Google Docs, a meeting gets scheduled in Google Calendar, and conducted remotely via Google Meet — with every single attendee apparently turning their camera off — and the whole thing is finalized with e-signatures. Cue the fireworks.

Since this is an ad from a tech company in 2026, AI has a role to play. The fictionalized founders use Google’s “help me visualize” AI tool to audition different animals for the national seal, Gemini takes notes on the meeting, and the founders consult the chatbot before declining King George III’s request to access the document.

The tone is tongue-in-cheek throughout — at one point Sam Adams asks, “Can we settle this over beers?” — and the AI evangelism is relatively restrained compared to many recent ads in the genre. Notably, the commercial stops short of suggesting that the Declaration of Independence would benefit from AI assistance, a line that has tripped up other brands. Perhaps the most AI-forward element is the footage itself, which carries the uncanny visual quality associated with AI-generated video.

Also Read: HubSpot’s Aja Frost on Marketing in the Age of AI Search

Viewer response has been mixed. Comments on YouTube and Instagram have been largely positive, but reactions on Bluesky have been sharply critical. Posters called the commercial “cringey” and “stunningly tone deaf,” with the AI angle drawing the most scrutiny — even as several users, including historian Angus Johnston, pointed out that it is “amazing how little of this is actually AI.”

“Even in a corny fantasy joke, it’s impossible to make the case that AI is a useful tool for political organizing, writing, or human collaboration,” Johnston wrote.

The post Google’s Fourth of July Ad Imagines AI at Independence Hall appeared first on MartechView.

Zeta Global, the AI Marketing Cloud, unveiled new findings from its latest AI shopping behaviour research, highlighting that consumers are increasingly willing to authorise AI agents to shop and buy on their behalf — signalling the early emergence of what the company calls agentic commerce.

Based on a survey of 2,000 U.S. adults who reported using AI to make a purchase within the past three months, the study marks the second instalment in Zeta Global’s AI shopping insights series. While fully autonomous shopping remains in its early stages, the findings suggest consumers are becoming more open to delegating portions of the purchase journey to AI-powered agents, with the strongest signals emerging among parents.

“There’s no question consumers are increasingly trusting AI with shopping decisions,” said David A. Steinberg, Co-Founder, Chairman, and CEO of Zeta Global. “The more important question now is whether brands are positioned to be discovered, recommended, and ultimately selected by AI. As consumers increasingly turn to AI to discover and evaluate products, brands need to know how they’re showing up.”

Also Read: HubSpot’s Aja Frost on Marketing in the Age of AI Search

Agentic Commerce May Change Discovery Before It Changes Transactions

While consumers are increasingly comfortable using AI to guide purchase decisions, they still prefer to complete transactions directly with brands. Seventy percent of AI shoppers said they would rather purchase directly from a brand’s website than buy through AI, suggesting AI is taking on a discovery and evaluation role while the actual transaction continues to flow through brand-owned channels.

Consumers also expressed a strong appetite for AI experiences built by brands themselves. Fifty-four percent of AI shoppers said they would choose a brand’s personalised AI experience over a general-purpose AI tool. Among consumers aged 18 to 45, that figure rises to 58%.

“When we conducted our first AI shopping study in late 2025, consumers were beginning to invite AI into their purchasing decisions,” said Pamela Lord, President of Customer Relationship Management at Zeta Global. “Just a few months later, we’re seeing signs that invitation is evolving into authorisation. As AI becomes a more influential layer in the purchase journey, brands need to understand how they show up in AI-driven recommendations and create experiences that are useful enough to earn the next click.”

Parents Are Leading the Shift

Parents with children under 18 are emerging as some of the earliest and most engaged adopters of AI-powered shopping experiences. Forty-three percent said they would allow AI to make purchases on their behalf within a set budget, compared with 27% of non-parents. The same proportion — 43% — said they would let AI automatically reorder household essentials, versus 31% of non-parents. Seventy-four percent said AI helped them discover a new brand they otherwise would not have considered, compared with 66% of non-parents, and 60% said they would choose a brand’s personalised AI experience over a general-purpose AI tool, versus 49% of non-parents.

Broader Shifts Across the Consumer Base

The survey also surfaced wider behavioural changes. Thirty-six percent of AI shoppers now spend less time researching purchases, while 59% say AI has reduced the likelihood they will return a purchase — a meaningful finding for retailers managing reverse logistics costs. Only 21% say they spend more money because of AI, and 29% say AI has made them less likely to shop in-store.

Also Read: Are We Using AI to Help Customers or Avoid Them?

AI Use Varies Sharply by Category and Demographic

Electronics is the leading category for AI-assisted shopping, with 33% of AI shoppers naming it their top AI category — rising to 45% among men. Among women, beauty products are the most common AI shopping category at 20%, compared with just 4% among men. Household items rank second overall at 21%, followed by clothing, jewellery, and accessories at 15%.

The survey was conducted online in May 2026 among 2,000 U.S. adults who reported using AI to make a purchase within the past three months.

The post Consumers Are Starting to Let AI Shop on Their Behalf: Zeta Global appeared first on MartechView.

In late 2022, before most marketers had worked out what to do with ChatGPT, Aja Frost was already making the case internally at HubSpot that large language models were about to change how people find software — and that HubSpot needed a strategy for it before the shift became obvious.

That pitch turned into a cross-functional initiative spanning growth, product, engineering, brand, and communications. The result: HubSpot became the most visible CRM in LLM responses, with citations up more than 4,000% and AI-referred demand up over 1,850%. It also led to HubSpot’s acquisition of Xfunnel and the launch of the HubSpot AEO Grader, the first free tool designed to help companies understand their visibility in AI answer engines.

Frost is now HubSpot’s Senior Director of Global Growth and Paid, leading the teams responsible for top-of-funnel demand through SEO, LLM optimization, and paid media. In this conversation, she talks about why the website has become the last stop — not the first — on the modern buyer journey, what it actually means to give algorithms more control without handing over strategy, and how a company that sells marketing software to marketers is thinking about AI automating a significant chunk of what those marketers do.

Excerpts from the interview; 

Your role sits at the intersection of growth, paid, and AI. What does your focus look like today?

My team owns top-of-funnel demand for HubSpot, with a focus on paid, SEO, and AEO. We sit at the intersection of Marketing, Sales, Analytics/Ops, and Product. Right now, a significant portion of my team’s time is on answer engine optimization (AEO) — making sure HubSpot shows up in the answers buyers are getting from ChatGPT, Gemini, and Perplexity — and turning what we’ve learned into a playbook our customers can use. Part of that work was recognizing there was no good way for marketers to see how the AI search landscape was shifting, so we built HubSpot’s AEO Sensor, a free tool that tracks AI visibility, citation, and traffic trends by industry. This helps people understand whether their strategy is working or the underlying models are changing.

Also Read: AI Ads Will Win Only If They Earn Consumer Trust

How has HubSpot’s growth playbook evolved from the inbound era to the age of AI search?

For years, inbound marketing was the playbook. It worked because buyers were searching on Google, and we met them there. Now, people are going to ChatGPT or Gemini, having an in-depth conversation where they identify a problem, evaluate solutions, make a shortlist, and then, and only then, go to your website. The website has gone from an early stop on the buyer’s journey to the final one. In response, HubSpot has stopped targeting high-volume educational keywords and started building visibility in answer engines. We’re getting less traffic, but it’s much more valuable: Customers who arrive after doing their research in an LLM convert at about 3x the rate of traditional search visitors. They’re pre-qualified and ready to purchase. 

How do you scale paid growth globally without losing local relevance?

When going global, marketers typically run a single playbook everywhere or fully decentralize, letting each region do its own thing. Neither works. We use a global operating system that includes shared segmentation, KPIs, and a shared narrative. Execution is local: we adapt channels, messaging, and offers to each market’s buying behavior and digital maturity. Lastly, we test before we expand. A campaign earns its way into new markets based on data, not assumptions about what should translate. 

Paid media has changed dramatically. How has HubSpot adapted, and what’s working today that wasn’t two years ago?

Paid got harder when cheap targeting stopped being reliable. iOS changes, cookie disruption, and rising CPMs have pushed the industry away from easy scale. We’ve had to get much more disciplined about where paid adds value — and much less dependent on third-party signals. First-party data and well-built intent signals are far more impactful now than broad reach. That has also changed what we optimize for. Traffic volume is no longer the North Star. We care more about visibility, branded demand, conversion rate, pipeline quality, and revenue because they are better indicators of whether we’re actually influencing buyers throughout a fragmented journey. 

In practice, that means paid is less about buying broad top-of-funnel traffic and more about amplifying strong signals and strong creative around higher-intent destinations. What’s working now that probably wouldn’t have worked two years ago is this combination of first-party precision, off-site amplification, and integrated paid support. The channels and content types that drive AI citations, such as YouTube, newsletters, podcasts, and Reddit forums, also happen to be where buyers spend time. Today, 90% of HubSpot’s leads come from non-blog sources, with YouTube leads up 100% and newsletter leads up 90%. The paid strategy follows the same logic. 

Also Read: The Agency-Led E-commerce Model Is Changing

How much control should marketers give AI-driven ad platforms—and where do you draw the line?

We’re giving algorithms more room than we used to, but we aren’t handing over the strategy. The most important parts are still in human hands: who we want to reach, what counts as real intent, what message we want in the market, and how we judge success. The machine can optimize delivery, but it shouldn’t define the objective.

Automation is worth it when it operates within strong guardrails and leverages first-party data, explicit intent signals, and strong creative. You’re giving up tactical control, not strategic control, and that’s only worth doing when the algorithm is paired with strong inputs and rigorous measurement. If you’re using automation with weak data or vague goals, it’s just spending efficiently against the wrong objective. 

Where has AI delivered the biggest measurable impact across HubSpot’s growth operations?

AI isn’t something my team does off to the side; it’s part of our day-to-day operating model. On the creative side, we use AI to generate creative and scale testing, produce search ad variations, and make sure assets are on-brand and speak to our persona before they go live.  Our internal heuristic is: use AI to go faster, but keep a human in the loop. We’re also using AI in optimization and execution, including value-based bidding, dynamic personalization, and A/B testing at scale. The results are pretty incredible: AI-referred demand is up 1,850%, email personalization has driven an 82% improvement in conversion rates, and 94% of HubSpotters use AI weekly.

Also Read: Payment Experience Is the Foundation of B2B Loyalty

As AI automates more marketing work, how is HubSpot redefining the marketer’s role?

AI is automating many traditional marketing activities, including content production, campaign setup, reporting, and personalization. But internally and across our customer base, we’re seeing demand for marketing “judgment” or taste rise.  AI raises the ceiling on what a marketing team can do — it doesn’t lower the floor on strategic thinking. At HubSpot, 94% of our team uses AI weekly. We’re not asking whether to use it; we’re asking whether the output is driving outcomes. Building on that, we don’t tell customers that AI replaces the marketer. We’re showing them how to use it to do more of the work that matters.

The post HubSpot’s Aja Frost on Marketing in the Age of AI Search appeared first on MartechView.

Talkdesk, a leader in Customer Experience Automation, announced Talkdesk Agent Builder, a natural language-driven tool that enables both business users and technical teams to oversee the complete lifecycle of their customer service AI agents. Customer experience leaders describe the outcome they want to achieve in plain language, and Agent Builder creates, tests, diagnoses, and validates the AI agent before it ever interacts with a customer. The result is a faster path to production, lower deployment risk, and the ability to build a trusted AI workforce in hours instead of weeks.

Building a production-ready AI agent has traditionally required weeks of iteration, testing, and tuning to keep it on topic. Agent Builder introduces a zero-prompt approach to AI agent development, compressing that process into hours and enabling organizations to rapidly build and scale a full AI workforce without sacrificing quality or control.

By functioning as a specialized agent-building agent, the platform automatically ingests an enterprise’s existing raw assets, such as standard operating procedures and policies, and translates them into high-precision logical instructions and structured guardrails. Once new AI agents are approved by a human operator and deployed, they are continuously monitored, evaluated, and governed through the CXA Operations Center. This complete framework helps organizations grow their AI workforce without multiplying their operational risk.

Also Read: AI Ads Will Win Only If They Earn Consumer Trust

The platform’s capabilities include plain-language agent creation, which allows operations teams to define an agent’s role, tone, behaviors, and edge cases using natural language, eliminating the need for prompt engineering or technical configuration. Pre-deployment validation automatically reviews instructions for consistency and completeness, surfaces gaps and ambiguities the team may have missed, and recommends specific actionable improvements before any agent reaches production. Simplified troubleshooting means that when an agent underperforms during testing or in production, Agent Builder diagnoses the issue, recommends fixes, and facilitates re-validation, with the highest-scoring version presented for human approval.

“You cannot simply deploy an AI agent into the wild and hope for the best,” said Munil Shah, Chief Product, Technology, and Customer Officer at Talkdesk. “Organizations need proof that an agent will follow instructions, stay on brand, and handle pressure before a real customer ever experiences it. Talkdesk Agent Builder provides that certainty. By systematically debugging and validating interactions against simulated datasets, we give enterprises the confidence to expand their operations with an AI workforce that’s trusted and reliable.”

Also Read: Your ERP Is Holding You Back. Here’s How to Fix It.

Talkdesk is showcasing Agent Builder and the broader Talkdesk Customer Experience Automation platform at Customer Contact Week Las Vegas, where Talkdesk CXA is a finalist for Automation Solution of the Year in the CCW Excellence Awards.

The post Talkdesk Launches Agent Builder to Speed AI Deployment appeared first on MartechView.