Every week, a new AI announcement promises to change everything. But what are the people actually running eCommerce and SaaS companies saying about it, away from the hype?
We went through dozens of podcast episodes, keynotes, and interviews to pull the most candid, insight-rich quotes from CEOs, founders, and business leaders who are navigating AI in the real world. Not what they are pitching to investors. What they are genuinely working through.
Here is what stood out, and what it means for businesses selling online in 2026.
1. The business case for AI: white space, not replacement
Tobi Lütke, Shopify’s CEO, has one of the more optimistic takes on what AI makes possible for entrepreneurs:
“People should have a product that the world wants. Ideally come up with a unique take on something. And there is so much white space out there. And I think AI should do everything else, so that if you show up with a product, you can start a business.”
— Tobi Lütke, CEO of Shopify | Uncapped with Jack Altman, Ep. 50
What Lütke is describing is a structural reduction in the operational overhead of starting and running a business. If you have a great product, AI lowers the barrier to handling marketing, support, operations, and content — everything that used to require significant headcount before you could scale. That is a genuine competitive shift, particularly for independent software vendors and early-stage SaaS companies competing against more established players.
2. Agentic commerce is the next big shift in how software and digital products get sold
Shopify President Harley Finkelstein is one of the most vocal voices on where digital commerce is heading, and his framing of “agentic commerce” is worth taking seriously:
“Since 2015 until now, we’ve had new channels open up, social commerce opened up. I think agentic commerce is this new front door which has the chance not only to take some market share from the existing modern retail/digital pie, but to actually convince more people that currently do not shop online, to do so. So, there is a TAM expansion opportunity around agentic commerce. I think it’s going to happen slowly and then all at once, where we’re going to use these agentic applications as personal shoppers. The coolest part about this is that it’s just another channel.”
— Harley Finkelstein, President of Shopify | 2026 Upfront Summit
The key phrase here is “just another channel.” For software and SaaS businesses, this has direct implications for how products get discovered, trialed, and purchased. AI agents are beginning to research, evaluate, and recommend software on behalf of buyers, which means the next acquisition channel may not be a search ad or a review site, but an AI assistant making a recommendation.
3. AI is a tool for your vision, not a replacement for it
Melanie Perkins, CEO of Canva, puts a boundary on AI enthusiasm that is worth internalizing:
“So, I really believe that AI should accelerate your vision and creativity, not override it. I think that it’s really important that AI is just another tool in our toolkit and it will help achieve our goals, if we choose to use it.”
— Melanie Perkins, CEO of Canva | The Verge / Decoder Podcast
For marketing, content, and product teams in SaaS and digital commerce, this is a useful guardrail. AI accelerates execution. The strategy, the brand perspective, and the taste still need to come from humans.
4. We are moving from a factory mindset to a lab mindset
Des Traynor, co-founder and CSO of Intercom, offers one of the most interesting reframes of how AI changes the way teams operate:
“What AI has introduced is this extreme layer of probabilism where you don’t know if something’s going to work. Because of that we are moving from this factory mindset into something that’s a lot closer to a lab mindset, which is experiments, uncertainty, empirical evidence, consistent evaluation, always trying new things. And every now and then you have a breakthrough, which needs to get commercialized. And in that world, you go back to the factory again where you’re like, we can do this, let’s build the software around it.”
— Des Traynor, Co-Founder & CSO of Intercom | OPTO Sessions
For SaaS teams, this is a genuine cultural shift. Shipping polished features on a fixed roadmap is giving way to shipping experiments quickly, learning, and iterating. The same applies to go-to-market: the winning teams in 2026 are not running the most perfect campaigns — they are running the most experiments per quarter.
5. The future is humans and agents working together
Yamini Rangan, CEO of HubSpot, describes where the practical endpoint of AI adoption lands for most businesses:
“Today, with AI, there is no map. So, you have to get comfortable with being an explorer. When we think about work, we think about what can be automated with AI, what can augment the productivity of teams, and how we can do things better. At HubSpot, we wanted to inspire the teams to learn so we can serve customers better. The future is hybrid. It’s going to be humans and agents in the loop, taking all the intelligence that agents and AI can deliver, making ourselves much more productive and doing higher value work.”
— Yamini Rangan, CEO of HubSpot | Silicon Valley Girl, Marina Mogilko
The “no map” framing is important. SaaS companies waiting for a defined best practice before investing in AI readiness are misreading the environment. The advantage right now goes to teams that are actively experimenting and learning, even when the outcomes are uncertain.
6. AI should remove the busy work so humans can do the high-value work
One of the most grounded takes on AI comes from Eric Glyman, co-founder of Ramp, who draws a sharp line between what AI can and cannot replace:
“I don’t believe that AI is smart enough to do the job of a CFO or a complete finance function, but it is definitely capable of doing your expense reports. Why not automate these terrible parts of your job away? It allows your best salespeople to go and spend that last hour selling and actually doing the work they were meant to do.”
— Eric Glyman, Founder of Ramp | Fortune
This is the most practical framing of AI ROI that gets missed in most conversations. The question for any SaaS or digital commerce team is not “can AI replace my department?” It is: “what are the low-value, high-volume tasks that AI can absorb so my team can do more of what actually drives growth?”
7. The next phase of marketing automation is fully autonomous
Andrew Bialecki, CEO of Klaviyo, describes where email and marketing automation is heading:
“We think that now it’s not enough to just have software you log into, and you, the user, have to know what you want, you have to do the work yourself. You have to point, and click, and design the creative, the rules. A lot more of that is going to be done autonomously. And autonomous for us means both the design, the creative design, but also the rule design like who to send it to, when, what’s performing, run the A/B test, is going to happen through AI and agents.”
— Andrew Bialecki, CEO of Klaviyo | Interview by Vlad Kachur
For SaaS and digital commerce businesses that rely on lifecycle email, onboarding sequences, trial conversion, renewal campaigns, and upsell flows, this signals where the major platforms are heading.
8. AI-generated content still needs a human edit pass
Hiten Shah, serial entrepreneur and investor, makes a point that most content teams need to hear:
“If you want your content to not sound like AI, don’t flag human brains that it sounds like AI. If you’re using AI to generate content, fine, we all are in one way or another. Make it sound like a human. Make it sound like you. Read it. Edit it. Change it. Who says you should take what it produces and throw it to the world and think that’s ok?”
— Hiten Shah, Entrepreneur & Investor | Best Story Wins Podcast
For SaaS and software brands investing in content marketing, email, and product copy, this is a workflow question, not just a quality question. The value of AI in content is speed and scale. The value of the human edit pass is brand voice, credibility, and differentiation — three things that matter enormously when you are trying to stand out in a crowded software category.
9. AI in customer support: get the backend right before going public-facing
Steve Chou, digital entrepreneur and educator, shares a pragmatic observation from his own experimentation with AI in customer service that translates directly to SaaS support teams:
“I’ve been experimenting lately with customer service bots. And this is where you train a bot to answer the most commonly asked questions. I actually do not like paying for third party tools. And what I found at least is that using the APIs that OpenAI and Claude give you, you can pretty much do it yourself. Coding is so much easier now with these tools. However, I found that people don’t like talking to chat bots. So, where I’ve been deploying them isn’t so much publicly, but more on the backend, with the customer service emails. You can have an AI know your most commonly asked questions are, and then draft a reply automatically. And you can have a customer rep that goes through and makes sure everything sounds right.”
— Steve Chou, Digital Entrepreneur | Sellernomics Podcast
The pattern here is directly applicable to SaaS support operations. Rather than replacing your support function with a public-facing chatbot — which users often push back on — using AI to draft responses to incoming tickets, which a human then reviews and sends, delivers real efficiency gains without the customer experience risk.
10. Real AI adoption in business is still in early stages
Eran Zinman, Co-CEO and Co-Founder of monday.com, is more measured about where AI adoption actually stands inside most businesses:
“Things are changing so quickly right now, it’s hard to keep up. And to see real AI usage in the real world, it’s still a process for a lot of SaaS vendors. I think a lot of companies try to build new features with AI. And there is a lot of friction; customers need to adopt new technology, they need to change how they use the product. People are using AI on a personal basis, but when it comes to their business, it’s still a big question mark how we’re going to implement it.”
— Eran Zinman, Co-CEO & Co-Founder of monday.com | SaaStr Podcast
This is a useful reality check for SaaS vendors building AI features. The gap between AI that works in a demo and AI that gets adopted consistently inside an organization is still wide. Friction reduction and change management matter as much as the technology itself. If your customers are not yet using the AI features you shipped, you are not alone — and the answer is usually in onboarding, not the model.
11. AI fails without a strong data foundation
Henry Schuck, CEO of ZoomInfo, makes one of the most important structural points in the whole AI conversation:
“Most often, these AI projects within companies, particularly in go-to-market organizations are failing because they don’t have the context for the company. There is not a data foundation that understands the business, understands their customers, knows every conversation they’ve had, knows what is happening in the businesses outside and brings that all in place, where AI can build on the full context of the business that is working on behalf of. And so foundational data is critically important.”
— Henry Schuck, CEO of ZoomInfo | Schwab Network
Before any SaaS or digital commerce business invests in AI tooling, the quality and accessibility of their customer data is the most important variable. Bad data in means bad outputs out, regardless of which model or tool you use. For subscription businesses in particular, clean data on customer behaviour, usage, and churn signals is the foundation everything else is built on.
12. AI scales your strategy, good or bad
Bryan Eisenberg, one of the original voices in conversion optimization, offers a caution that every team chasing AI-driven scale should sit with:
“AI can write clever emails. It can test headlines faster than your team can finish a meeting. It can even mimic your brand tone across platforms. But it can’t fix a broken story. If your departments aren’t aligned, AI won’t fix that. It will just spread misalignment more efficiently.”
— Bryan Eisenberg | bryaneisenberg.com
For SaaS businesses where sales, marketing, and product often operate in silos, this is the organizational warning that precedes the AI investment conversation. Before scaling output with AI, the strategic and internal alignment work needs to be right. AI is a force multiplier; if the underlying strategy is off, AI multiplies the problem.
13. Experience is an advantage in the AI era
Ethan Mollick, AI researcher and professor at Wharton, makes a counterintuitive point about who benefits most from AI:
“If Claude is really good running your company, Claude is also good at running every other company, and there’s no variation between them. And generically, high quality with no variation means there’s no competitive edge. I think humans who bring competitive edge to this, one way or another, just by providing variation if nothing else, is a useful way to think about problems. The more experienced you are, the better you’re going to be at using AI, if you decide to use it.”
— Ethan Mollick, AI Expert & Wharton Professor | A Bit of Optimism Podcast with Simon Sinek
This reframes the AI threat narrative entirely. AI democratizes access to competence. But experienced operators — people who deeply understand their market, their customers, and their product — will use AI better than those who are new to the domain. In software and SaaS, where domain expertise and customer insight compound over years, that experience becomes more valuable, not less.
14. Bottom-up AI adoption is not enough: leadership needs to lead
Andrew Ng, one of the most respected voices in AI, challenges the “let a thousand flowers bloom” approach to enterprise AI:
“All of us have invested in the bottom-up innovation, the ‘let a thousand flowers bloom’ strategy, and for the most part is not paying off. So, CEOs and boards are asking, where is the ROI for AI? I think we should keep on investing in bottom-up innovation, let’s keep on doing that. But it turns out that bottom-up innovation often results in point solutions that drive incremental efficiency gains, which are actually a good thing, but not the transformation, not the broader transformation that AI has been promising us. So, bottom-up innovation is really good, it generates a lot of ideas, and that has to be complemented with a top-down motion of having someone with the broader scope to change how all of these steps operate, to then create growth.”
— Andrew Ng | Interrupt 26 Event
This is the governance question most SaaS leadership teams are now wrestling with. Individual teams experimenting with AI is necessary but not sufficient. Realizing the larger opportunity — whether that is faster product development, reduced support costs, or more efficient customer acquisition — requires someone with cross-functional scope making deliberate decisions about how AI changes operations end-to-end.
15. Leaders need to use AI themselves to make good calls about it
Aaron Levie, CEO of Box, makes a point that applies to every C-suite that is trying to guide AI strategy from a distance:
“A CEO, by definition, is the furthest away from the real work that’s happening in the company. And it’s very easy to be like, ‘I can just automate that engineer’ or ‘I can go automate that marketing campaign’. But when you’re closer to the problem, you realize that you probably can’t just have an agent go and do all that without any human supervision because it’s going to do the wrong thing, or the taste of what the agent is going to deliver is going to be off, or it’s going to introduce a bug. So that keeps humans in the loop for the foreseeable future. So, the thing that I infer to CEOs is to use the technology actually so much, that you get to the other end of that psychosis and you can actually see in a much more practical and pragmatic way all the places where humans are still necessary to get the gains from this technology.”
— Aaron Levie, CEO of Box | CXOTalk Podcast
This is the leadership accountability point. You cannot develop a credible AI strategy for your business if you are not actively using the tools. The judgment required to make good decisions about AI delegation only comes from direct experience with where it works and where it breaks.
What the patterns tell us
Across all of these conversations, a few consistent themes emerge:
- AI is a new acquisition and engagement surface, not just an efficiency tool. Agentic commerce is creating new ways for software and digital products to get discovered and purchased. The businesses building visibility in AI-driven recommendation environments now will have an advantage that compounds.
- Data quality is the foundation. Almost every practical failure story in enterprise AI traces back to poor or fragmented data. Before adding AI tooling, audit what your customer and product data actually looks like — especially if you are a subscription business relying on behavioral signals for retention and expansion.
- Human judgment is more valuable, not less. Experience, domain knowledge, and strategic thinking are not made redundant by AI. They are what make AI useful. The experienced SaaS operator who uses AI tools well will consistently outperform the AI operating without that context.
- Bottom-up experimentation needs top-down direction. Individual teams using AI to speed up their existing work is valuable but limited. Realizing larger gains — reduced CAC, faster time-to-value, lower churn — requires deliberate, cross-functional decisions about how AI changes how the business operates end-to-end.
- Speed to learning matters more than speed to perfection. The shift from factory to lab mindset, as Des Traynor puts it, is real. The SaaS teams building the capacity to run experiments quickly, learn, and adjust will compound their advantage over time.
One more thing: your commerce infrastructure needs to keep up too
If AI is changing how software gets discovered, evaluated, and bought, it follows that the infrastructure behind how you sell and get paid needs to match that pace.
If you are rethinking how your business is set up to sell in an AI-driven environment, see how 2Checkout works or talk to our team.
