Forget The Chip Numbers. AI's Next Productivity Wave Is On The Retail Floor

This week Broadcom (AVGO) raised its AI revenue guidance again to $58 billion, up 186% year over year. It's the kind of forecast that commands a news cycle, and it's also exactly what investors have been trained to look for when they ask what AI is doing for productivity. 

The answer, in many cases, continues to be routed through the same handful of categories including chip demand and coding assistants. To be clear, those are real and they're growing. 

At the same time, they're also the first place to look, and these in many cases are rarely where the next leg of investment returns comes from.

Across global tech markets, the pattern that keeps repeating is that the biggest productivity gains show up first in the markets that people aren't watching, precisely because those have the most inefficiency left to wring out. 

Software has already eaten the easier parts. What's left is physical, in-person, and also voice, and that's exactly where AI is starting to move fastest.

Here are three markets for investors where that shift is already showing up in the numbers.

Physical retail is still the biggest pool of commerce on earth, and it's just now getting its AI layer.

It's easy to forget how dominant in-person retail remains, in part because so much media coverage is built around e-commerce growth curves. 

For instance, Forrester's most recent forecast puts total US retail sales at $5.2 trillion in 2025, climbing to $6.2 trillion by 2030. What’s surprising to many is that in-store sales are expected to still account for 71% of that total.

E-commerce, for all the coverage it gets, tops out around 29%. Investors who have spent the last decade positioning entirely around digital commerce have been correctly riding a real trend, but they've been ignoring the much larger pool sitting right next to it.

That pool has been notoriously hard to instrument. A retailer can know everything about a website visitor's behavior, and at the same time almost nothing about what's actually happening on a physical shelf three states away, including whether the product is in stock.

This is the gap agentic AI is beginning to close. Effie.ai, referred to as the first agentic retail execution platform, uses AI image recognition to analyze shelf conditions in real time and generate store-specific action plans for field teams, closing the loop between headquarters and the store floor automatically instead of through manual reporting. 

The company, based in Chicago, already works with major brands like Nestlé. That's not a hypothetical use case, it's an indication that the physical retail layer, long considered too fragmented and too manual to digitize, is becoming investable the same way e-commerce infrastructure was a decade ago.

According to Manu Swami, Chief Technology Officer at Sonata Software, “AI is no longer just another technology layer. It is becoming the operating intelligence of the enterprise.” We should expect physical retailers to quickly realize this as well. 

Publicly traded companies to watch here include Walmart (WMT) and Kroger (KR), which uses smart digital shelf technology called Kroger Edge and predictive ordering to cut down on food waste.

Voice AI is quietly becoming the highest ROI purchase a founder or small business owner will make

When it comes to the growth of AI, it's hard for investors to miss the voice AI opportunity in front of us. Upfirst AI is a good example of how quickly this is scaling. Co-founded by Alfredo Salkeld and Gene Sigalov, who founded and led SimpleTexting to millions in annual revenue, Upfirst built an AI receptionist for small businesses that’s reaching new milestones each month. 

According to Upfirst, the value at risk for a single unanswered call sits at around $1,900 for a roofing company and around $620 for a legal practice. Voice AI can clearly play an important role here.

Anjli Jain, General Partner at ElevenX Capital, added that voice AI is no longer a future concept; it is rapidly becoming the preferred interface between people and intelligent systems. Organizations that combine conversational experiences with trustworthy AI, enterprise integration, and operational excellence will be best positioned to lead the next generation of digital transformation.

We should also expect to see a rise in solo founders from the rise of these solutions, with the data backing this up. 

According to Carta, solo-founded startups made up just 23.7% of new U.S. companies in 2019. By the first half of 2025, that number had jumped to 36.3%, more than one in three. 

Said Ibrahim Hasanov, Founder of MyUser, “Most people who talk about solo founders talk about grit, like the whole problem is a mindset issue. It's not. The problem is math. One person has a finite number of hours, and a company needs someone doing sales, someone doing support, another writing code, most often at the same time. For most of startup history, that math simply didn't work past a certain size. You hired, or you stalled. That math has changed. Not because AI makes founders lazier or smarter; rather because of the number of hats you have to physically wear yourself.”

It also isn’t only voice AI helping in new business creation here: Global workforce platforms like Ontop, which allow companies to better pay remote workers, are playing an important part for new businesses. Additionally, AI coaches like Cloverleaf that fully understands a company's people and the context of the workday, are must-watch.

Publicly traded companies to watch here include SoundHound (SOUN) and Cerence (CRNC).

Blue-collar work is where AI adoption is happening the fastest relative to how little capital has flowed into it.

If you want to find a market where AI productivity gains are large and underpriced, look at where the labor shortage is most acute. The U.S. has struggled for years to fill skilled trade positions, and that gap has forced more pragmatic AI adoption in blue-collar sectors than in categories that get more investor attention.

Vic Pellicano, CEO of SkyPSI, has built his business with the belief that AI-guided drones and automation in commercial cleaning aren't novelties, they're becoming the standard way smaller operators compete for large scale contracts that used to require far bigger crews. Pellicano has previously built and sold multiple companies, and he talks about this shift the way an operator does rather than the way a technologist does: the value isn't the algorithm, it's the person running the drone who can now create a better future for themself.

According to Pellicano, “No one is writing the version of the story where AI hands a plumber a business he couldn't have run five years ago, or lets a two-person cleaning outfit compete for a contract that used to require thirty employees and a call center.”

It's worth noting as well that infrastructure, which often doesn’t make the news, is likely where the competitive advantages are being furthered the most. HIPAA-ready white-label telehealth platforms like QuickBlox, which provide the chat, voice and video infrastructure are part of why deploying AI into a small clinic no longer requires a custom engineering team. The same is true of solutions like Chromatics AI, which is giving modern brands AI infrastructure.

None of the markets above will necessarily show up on the first page for investors. They don't have the brand recognition of the large cap names like Anthropic dominating this year's AI conversation. However, productivity gains follow inefficiency, and this has always been highest in the markets that got skipped the first time software went looking for problems to solve. 

Physical retail, voice-driven small business operations, and blue-collar trades were skipped for a reason, including that they were harder to reach and sell into. AI is finally becoming efficient enough to reach them anyway. 

For investors willing to look past the obvious categories, that's where the next round of productivity-driven returns is most likely to be underpriced.