What’s the BUZZ? — AI in Business

Your next customer may not be a person. When a personal assistant books the flight and fills the cart, the last click belongs to an agent, and the ad you served a human on social media no longer shows up as the sale. So where does advertising go when the audience is an AI agent?
In this episode of "What's the BUZZ?", host Andreas Welsch sits down with Walter Schmidli, Leader of Paid Social Media, to work through what agent-led shopping means for paid media. Walter focuses on paid social for niche industries and on how AI is reshaping digital marketing.
Key insights from this conversation include:
Advertising does not go away. It moves up the funnel. Walter expects the last click to be executed by an agent, but people will still browse social media and watch TV. That is where ads reach them, and reach and frequency still drive conversion.
The impulse purchase is under pressure. Measurement gets harder too. Walter says the age of buying on the fly because the right ad reached you at the right time may be coming to an end, and conversion may take longer to show up.
Start with SEO and GEO. Walter calls it the key short-term step for brands: make sure your SEO and GEO strategies are on point and tuned for agents. Information that is available and crawlable is what influences them.
Agents are not moved by emotion. People are. Walter says the descriptive detail has to live on a webpage or social platform that agents and LLMs can scrape, while the emotional side of creative is where people come in.
Social platforms feed the models, selectively. Walter says LLMs have become more precise about sources. It is not an average LinkedIn post that gets cited, but official spokespeople and key opinion leader content. Brands should find ways to get people to publish on those platforms, and test what works.
Scale will set the default. Walter expects an arms race over who rolls out agents first, and open questions about whether every brand gets its own agent and how Meta, Amazon and Google share data. Sponsored results and integrations may become a new way to monetize.
What leaders can do now: make all of your data available in a form agents and LLMs can digest, with SEO and GEO current as the baseline; shift ad messaging from buy right now toward consider it in the future; and keep watching the growing mix of publishers and agents to find the best fit.
Walter's line to write down: "Advertising has its place potentially just a little further up in the funnel."
If an agent makes the final click, where in your funnel does your advertising need to work harder?
Questions or suggestions? Send me a Text Message.
Support the show
***********
Disclaimer: Views are the participants’ own and do not represent those of any participant’s past, present, or future employers. Participation in this event is independent of any potential business relationship (past, present, or future) between the participants or between their employers.

Level up your AI Leadership game with the AI Leadership Handbook (https://www.aileadershiphandbook.com) and shape the next generation of AI-ready teams with The HUMAN Agentic AI Edge (https://www.humanagenticaiedge.com).

More details:
https://www.intelligence-briefing.com
All episodes:
https://www.intelligence-briefing.com/podcast
Get a weekly thought-provoking post in your inbox:
https://www.intelligence-briefing.com/newsletter

Your AI agent is about to be handed your credit card. Today, it recommends products. Soon it will complete the purchase and pay for it for you. Who is responsible when it buys the wrong thing, and will consumers trust it enough to let it?
In this episode of "What's the BUZZ?", host Andreas Welsch sits down with Michael Nevski, Consumer Payments Market Intelligence Expert, to work through what happens when AI agents start making financial transactions on our behalf. Michael draws on more than 15 years of consumer research in financial services and payments.
Key insights from this conversation include:
Agentic commerce is coming, but it is mostly hypothetical today. Michael defines agentic AI as systems that act autonomously toward a goal, unlike generative AI, which produces outputs that still need a human to act. Payment networks have started preparing, but privacy, security, chargebacks, and disputes have to be solved before a true network is enabled.
Consumers are excited and afraid at the same time. In Michael's research, 77% expect agentic commerce and generative AI to improve their lives, yet 83% have concerns about oversight and reliability. He calls it high hope and high fear, and says the fear comes from not understanding the black box behind the purchase.
Early trust has a small price tag. Asked how much they would let an agent spend, people said no more than $ 10 to $ 20. Michael compares it to play money in Vegas, an amount you are ready to lose.
Nobody agrees on who is responsible. Michael says the card issuer should still authorize the payment. When a transaction goes wrong, 39% of consumers point to the agentic AI service provider, 17% to the AI developer, 14% to themselves, and 6% to the payment networks. They want every party in the ecosystem to have a defined role.
Data sharing depends on the task. Only 7% would share location data in general, but that jumps to 45% when it helps find a restaurant for a business lunch within an hour, and browsing data goes from 8% to 43%. Consumers will trade data for a specific outcome, but as Michael puts it, "I don't want you to be Big Brother."
Preparing is everyone's job. Michael's list covers understanding early adopters such as Gen Z and millennials, merchants making product information readable for agents, issuers enabling tokenization, and working with regulators. Andreas adds that IT, finance, marketing, and supply chain all need to engage now.
What leaders can do now: make your product and service data clear and structured so an agent can read, compare, and verify it; give customers simple opt-in and opt-out controls over what an agent may do; review your legacy systems and work with partners and regulators on who is accountable when a transaction goes wrong.
Michael's line to write down: "Prove it to me that I can rely on that system."
If an agent made a purchase for your customers tomorrow and it went wrong, would they know who to turn to, and would they still trust you?
Questions or suggestions? Send me a Text Message.
Support the show
***********
Disclaimer: Views are the participants’ own and do not represent those of any participant’s past, present, or future employers. Participation in this event is independent of any potential business relationship (past, present, or future) between the participants or between their employers.

Level up your AI Leadership game with the AI Leadership Handbook (https://www.aileadershiphandbook.com) and shape the next generation of AI-ready teams with The HUMAN Agentic AI Edge (https://www.humanagenticaiedge.com).

More details:
https://www.intelligence-briefing.com
All episodes:
https://www.intelligence-briefing.com/podcast
Get a weekly thought-provoking post in your inbox:
https://www.intelligence-briefing.com/newsletter

Your next customer may never see your ad. More and more shoppers ask an AI assistant for the best option and take the answer it gives them, and agents are starting to recommend and buy on our behalf. So how do you advertise to an AI agent?
In this episode of "What's the BUZZ?", host Andreas Welsch sits down with Maarten Albarda, Marketing Executive and Consultant, to explore how marketing must change when AI agents shop, recommend, and buy for us. Maarten focuses on the "how" of marketing: the process, structure, agency ecosystems, and capabilities behind good marketing.
Key insights from this conversation include:
The time to learn is now, but the payback is not guaranteed. What you learn today may not matter tomorrow. Maarten argues it is still far easier to learn now than to unlearn and relearn five years from now, and it makes you a better user of AI either way.
Judge every AI change by its outcome. Maarten's test is that the result is at least the same as before, and if it is not better, it should be cheaper or faster. He hears a lot of AI anxiety from CMOs and CEOs and says the right pace is the one your organization can absorb, because "it's quite easy to break an organization with too much too soon."
The savings have not reached advertisers yet. Agencies have always sold human time, and agents that plan and optimize media mean fewer people and fewer hours. Maarten says clients aren't seeing that in their agency contracts yet, and advertisers are starting to ask what outcome they are buying instead of paying for the process.
Hourly-rate businesses change fastest. Maarten sees the same shift in legal, consulting, and financial advice. Services built on selling hours will feel the biggest impact over the next five years.
AI search may be hurting traditional search more than reported. Each platform reports only on itself, so independent numbers are scarce. Maarten expects the impact is bigger than we are told, because most people see an AI-generated answer first, often on their phone, and may go straight to a retailer app.
Find out how AI sees your brand. Ask the same question in GPT, Gemini and Claude and the answers and their sources differ. Maarten's first step is to buy a tool that shows which sources each model values for your category, then work out how to show up in that conversation.
What leaders can do now: link every AI initiative to a business purpose and measure a before and after, so you can show the outcome is the same or better; use a tool to find the sources GPT, Gemini and Claude rely on for your category, then plan how to show up in them; ask your agencies how AI savings are passed on, and move toward paying for outcomes.
Maarten's line to write down: "don't do AI for AI's sake."
If an agent chose between your brand and a competitor tomorrow, would you know which sources it is listening to?
Questions or suggestions? Send me a Text Message.
Support the show
***********
Disclaimer: Views are the participants’ own and do not represent those of any participant’s past, present, or future employers. Participation in this event is independent of any potential business relationship (past, present, or future) between the participants or between their employers.

Level up your AI Leadership game with the AI Leadership Handbook (https://www.aileadershiphandbook.com) and shape the next generation of AI-ready teams with The HUMAN Agentic AI Edge (https://www.humanagenticaiedge.com).

More details:
https://www.intelligence-briefing.com
All episodes:
https://www.intelligence-briefing.com/podcast
Get a weekly thought-provoking post in your inbox:
https://www.intelligence-briefing.com/newsletter

AI adoption is often treated as a one-size-fits-all approach. But individual employees and users have individual needs and preferences that successful AI programs consider.

In the latest episode of "What's the BUZZ?", host Andreas Welsch sits down with Christina Fischer, principal AI consultant and founder of Human Pattern Works, to unpack why so many AI transformations quietly fail the people meant to carry them, and what an actually inclusive AI playbook looks like.

Here's what you'll walk away with:
1) The real reason most AI initiatives fail
Christina puts the number at around 70%, and by her account, the common thread isn't the technology. It's leaders running a business problem like a software rollout, skipping the work across process, culture, and compliance that decides whether any of it sticks.

2) Your edge cases are your early-warning system
The employee who needs clarity on objectives or flexibility in how they work is often the first to feel friction nobody else has noticed yet. Listen to them early, Christina argues, and you catch what's breaking before it costs you the rest of the team.

3) Trust beats surveillance, every time
Keystroke logging and return-to-office mandates measure the easy thing, not the valuable one. Christina expects the flexibility of the pandemic years to come back, because it's what lets companies keep the specialized people AI transformation actually depends on.

Actionable takeaways you can implement this week
- Score before you scale: run Christina's four-dimension audit,  process, work culture, compliance, technology, on your own team before you greenlight another AI pilot.
- Build one real feedback loop: swap a status update for an open conversation about what's actually working, and treat what you hear as signal, not noise.
- Check on your middle managers: ask the people absorbing pressure from both directions how they're doing, before they become your next regretted departure.

Who on your team is the canary in the coal mine right now, and are you actually listening to them?
Questions or suggestions? Send me a Text Message.
Support the show
***********
Disclaimer: Views are the participants’ own and do not represent those of any participant’s past, present, or future employers. Participation in this event is independent of any potential business relationship (past, present, or future) between the participants or between their employers.

Level up your AI Leadership game with the AI Leadership Handbook (https://www.aileadershiphandbook.com) and shape the next generation of AI-ready teams with The HUMAN Agentic AI Edge (https://www.humanagenticaiedge.com).

More details:
https://www.intelligence-briefing.com
All episodes:
https://www.intelligence-briefing.com/podcast
Get a weekly thought-provoking post in your inbox:
https://www.intelligence-briefing.com/newsletter

Your next customer might not be a person. It might be an agent that searches, compares, and negotiates on someone else's behalf. So what happens to advertising, to pricing, and to trust when that becomes normal?
In this episode of "What's the BUZZ?", host Andreas Welsch sits down with Michael Carroll, Strategic Advisor and Fellow at LNS Research, to work through what actually changes when the customer is an AI agent. Michael spends his time on advisory boards and AI research, focused on what this technology should do for people, not only on what it can do.
Key insights from this conversation include:
The web was built for clicks. Agents are built for intent. Michael frames the shift as moving from a click-based architecture and economy to one organized around what you want. Today you do most of the orchestration yourself to get it. Reasoning at the edge flips that: the agent handles the orchestration, and the effort left to you is working out why you want what you want.
Ads do not disappear. They change target. Michael expects ads to keep running on TV and on your phone, but their job changes. Instead of persuading you to buy, they work to bias you into telling your agent what your preference is. The emotional pull advertising relies on does not land on an agent, so it gets aimed one step earlier, at you.
Intermediation will go autonomous. Mediation needs permission. Michael splits organizational work into intermediation, the moving and translating of information, and mediation, changing something from one state to another in a way that requires approval. The first is largely reversible and a strong candidate for autonomy. The second carries responsibility, and a machine never bears responsibility.
Causal reasoning is the trust architecture, not explainability. His verdict on the alternative is blunt: "Explainable AI is an abomination. It's an invitation to re-litigation." A causal chain of reasoning, he argues, is auditable and defensible, and it is not the same thing as deterministic.
Whoever closes the triad holds the influence. The reasoning agent at the edge, the platform you view the world through, and the architecture of trust. Close all three and you have something controllable, which raises the question of who is positioned to do it.
The collapse of distance points back at the individual. Michael cites Stanley Milgram's six degrees of separation from the 1960s and LinkedIn's own figure of roughly three today, and argues agents take it to one. His read is that this pulls people closer to everything rather than further apart.
What leaders can do now: make your business legible to agents and not only to people, with structure, schema, and clear answers to the questions an agent will ask on a buyer's behalf; sort your processes into intermediation and mediation, treating the first as candidates for autonomy and the second as needing a permission and audit trail before an agent goes near them; and work out, for a vendor agent and a customer agent transacting with each other, what each side is authorized to do, what information it is acting on, and whether it can be swayed.
Michael's line to write down: "If you can't trust it, it won't happen."
If your customer's agent evaluated you today, what would it find?
Questions or suggestions? Send me a Text Message.
Support the show
***********
Disclaimer: Views are the participants’ own and do not represent those of any participant’s past, present, or future employers. Participation in this event is independent of any potential business relationship (past, present, or future) between the participants or between their employers.

Level up your AI Leadership game with the AI Leadership Handbook (https://www.aileadershiphandbook.com) and shape the next generation of AI-ready teams with The HUMAN Agentic AI Edge (https://www.humanagenticaiedge.com).

More details:
https://www.intelligence-briefing.com
All episodes:
https://www.intelligence-briefing.com/podcast
Get a weekly thought-provoking post in your inbox:
https://www.intelligence-briefing.com/newsletter