Big Lessons From Small Players: What Retail Executives Can Learn From SMB AI Adoption
Artificial intelligence is evolving quickly in retail, but adoption looks very different depending on the size of the company. Large retailers have resources and scale, but smaller businesses are often the ones experimenting faster and closer to the customer. That contrast offers lessons for executives at companies of all sizes.
Different Scales, Different Approaches
Large retailers such as Walmart and Target are investing heavily in AI for revenue generation, inventory management, and competitive intelligence. They look to AI to make decisions across global operations, from optimizing store productivity to analyzing market trends.
Midsized retailers, with dozens or hundreds of employees and regional footprints, use AI to drive operational efficiency. Their focus is less on differentiating products and more on integrating AI into existing systems to streamline marketing, supply chain, and store management.
Small businesses, however, take a more personal path. Often with only a few employees and deep ties to their communities, they use AI tools to fill staffing gaps, manage leads, and keep customer relationships strong. Owners may literally text with customers or rely on AI to make sure no call or inquiry slips through. For these retailers, AI isn’t about scaling globally; it’s about sustaining local trust while protecting the bottom line.
Across all three tiers, the common thread is trust. Whether through personal relationships, operational reliability, or brand consistency, retailers live or die by the confidence customers place in them.
Promises vs. Reality
The shift to e-commerce offers a useful parallel. In the early 2000s, countless providers promised “one-click” solutions that would move retailers online overnight. Most of those companies failed. Their tools couldn’t deliver on the complexities of retail — e.g., managing inventory, monitoring customer reviews, tracking sentiment, and providing the same quality of service shoppers expected in-store. Today’s AI tools present a similar dilemma. They may look sleek on the surface, but without accuracy and integration, they risk undermining service quality. And in retail, poor service doesn’t just disappoint; it can permanently cost customers.
Risks of Off-the-Shelf Models
Public large language models (LLMs) add another layer of risk. They often require sharing sensitive data, lack the context of a retailer’s unique operations, and create silos when different departments deploy disconnected tools.
Walmart’s recent experiment with dozens of AI agents illustrates the problem. Each agent handled a small task, but none could “speak” to the others. In an industry where decisions about shelves, shipments, and supply are tightly interconnected, siloed agents replicate inefficiencies rather than eliminate them. This promptly triggered Walmart to change its own strategy as things started to get too confusing.
Why Data Ownership Matters
Furthermore, retail executives increasingly see that AI adoption isn't just about new features. Rather, it's about controlling the data behind them. Customer histories, pricing strategies, and supply chain information are too sensitive to risk in public models.
The lesson echoes the cloud era. Retailers initially moved workloads to the cloud for scalability, but with AI, many are reversing course, bringing workloads back on-premise to reduce costs and safeguard data. On-prem deployments address two key issues: privacy and unit economics.
Lessons From SMBs
This is where SMBs provide an important blueprint. With lean resources, they cannot afford cumbersome systems or privacy missteps. Instead, they adopt narrowly focused tools, test them quickly, and refine them based on direct customer feedback. Their agility allows them to balance efficiency with personalization and preserve the trust that keeps customers coming back.
Larger retailers can learn from this approach. Rather than rushing to adopt broad, generic platforms, they can emulate SMBs’ agility and specificity, deploying AI where it supports operations most directly, while keeping trust and accuracy at the center.
For example, a local boutique might use AI to automatically follow up with customers by text when new products arrive, ensuring they feel recognized and valued. A larger apparel chain could borrow this playbook by deploying AI to personalize regional promotions or tailor inventory recommendations based on local buying habits. In both cases, the technology works best when it's narrow, accurate and deeply connected to customer relationships — not when it's spread thin across dozens of disconnected systems.
The Takeaway
For executives across the retail industry, the message is consistent: AI success won’t come from generic, one-size-fits-all solutions. It will come from adopting systems that enhance accuracy, protect data, and reinforce customer trust. SMBs are showing the way, and the retailers that learn from their example will be best positioned to thrive.
Suman Kanuganti is the CEO of Personal AI, a small language model platform engineered for scaled experiences that are private, programmable, and precise.
Related story: The AI Advantage: 3 Ways AI is Reshaping Checkout Experiences for SMBs
Suman Kanuganti, based in San Francisco, is an AI entrepreneur who has been pioneering practical AI applications for over a decade. As Co-Founder and CEO of Personal AI since 2020, he has worked with large telco and retail partners globally, innovating how enterprises deploy AI at scale, including new revenue generating products, internal productivity use cases, and serving the large SMB market. Prior to Personal AI, Suman founded Aira in 2015, building it into a $50M revenue AI-powered accessibility company that transformed business experiences for people who are blind or have low vision. Suman regularly shares insights at premier events, including TechCrunch Disrupt, WSJD Live, Qualcomm Snapdragon Summit, CES, and SXSW. He continues to drive the future of enterprise AI through precision, scalable unit economics, and distributed infrastructure.





