Forrester research reveals both the opportunity and the roadmap for retailers ready to lead in agentic AI with governance and trust at the foundation.

Attentive sat down with guest speaker Sucharita Kodali, VP and Principal Analyst at Forrester Research, to understand what agentic AI means for retail. New Forrester research reveals that one in five US consumers trust AI agents with autonomous purchasing. With the right approach to governance and compliance, retailers can leverage agentic AI across operations, customer experience, and marketing today while building for broader adoption tomorrow.

‍

Thanks for having this conversation with us, Sucharita! Let's start with the basics: What does agentic AI mean?

To understand agentic AI, we need to step back. For most of AI's history, we've dealt with analytical or predictive AI (recommendation engines, fraud detection). Then generative AI arrived with ChatGPT in late 2022, and suddenly everyone was using large language models. But it's still assistive. You prompt, it generates, a human decides.

Now we're entering the agentic AI era. Agentic AI refers to systems where software entities perceive their environment, make decisions, and independently take actions to achieve goals without constant human oversight. The key word is "independently."

Forrester defines agentic AI as systems that enable agents to operate with increasing levels of autonomy. Here's what's critical: Automation isn't synonymous with autonomy. Traditional automation follows if-then rules. Agentic AI adapts, makes judgment calls, and adjusts based on context.

‍

So what does this look like in retail?

In retail, this plays out across three major areas: operations, customer experience, and marketing.

On the operations side, agents manage inventory, adjust pricing dynamically, handle supplier relationships, and optimize fulfillment. Amazon's Project Amelia is a good example.

For customer experience, agents handle returns, process exchanges, answer product questions, and eventually make purchases on behalf of shoppers.

In marketing and merchandising, agents create campaigns, personalize content at scale, optimize product placement, and adjust promotional strategies in real time. This is where many retailers are seeing immediate value because the applications are more contained and easier to govern.

‍

This sounds transformative. But based on your research into the evolution of these technologies, are we actually at the point where agentic AI is working in retail today?

Let me be very clear: It's so early stage. I can't stress that enough.

I talk about a critical maturity model: Automation leads to AI-powered automation, which eventually leads to agentic AI. Agentic AI will happen, but automation needs to lead the way. You can't skip steps.

My colleagues at Forrester have mapped the evolution in stages. Short-term, we're seeing agents that pull from your company's data and knowledge base to enhance their responses, but they're still following pretty strict guidelines.

Next, agents that can follow predefined workflows. Think of a return agent that validates eligibility, checks inventory, processes the refund, and updates customer records, as long as it's following a clear workflow you've established.

Later, agents that collaborate. An inventory agent talks to a pricing agent, which coordinates with a marketing agent to run a promotion on overstocked items. Working together without hand-offs.

Long-term, years not months, agents controlling complex strategic processes independently. But we're not there yet. These early iterations aren't truly autonomous. They're helpful, but they're not making independent decisions.

‍

Where are we seeing this work in retail today?

The most successful applications of agentic AI in retail are in contained environments where you can test and scale gradually.

Customer service agents handle routine inquiries, straightforward returns, and product questions. When something gets complex, they escalate to humans.

Inventory and pricing optimization is delivering real value. Agents monitor stock levels, adjust prices based on demand, and coordinate replenishment.

Content creation and merchandising is another sweet spot. Product descriptions, marketing copy, personalized recommendations, dynamic product placement. Retailers are producing content at unprecedented scale.

Where it's not ready yet: High-stakes purchasing on behalf of customers. Complex product advisory requiring deep expertise. Sensitive customer situations requiring empathy. Strategic business decisions defining brand positioning.

‍

What does this shift to agentic AI mean specifically for retail marketing teams?

Marketing teams are in a strong position because many agentic applications are ready today and deliver measurable ROI.

Performance marketing is seeing real results. Agents automatically reallocate budgets across channels, run A/B tests at scale, optimize send times, and adjust targeting. These capabilities are available now.

Content operations is transforming. The volume of personalized content modern retail requires makes this a natural fit. Marketing teams are producing content at unprecedented scale.

Customer engagement is being reimagined. Personalized email journeys that adapt based on behavior, dynamic recommendations that learn and improve, triggered campaigns based on real-time signals.

The key is starting with high-volume, lower-risk marketing tasks and building governance frameworks. Marketing teams that invest in proper controls now will be positioned to scale as technology matures.

‍

Recent Forrester research reveals interesting insights about consumer readiness. What did you find?

About 20% of U.S. online adults trust AI agents with autonomous purchasing. That's a significant early adopter segment.

The demographic breakdown is revealing. Roughly 26% of men and 25% of Gen Zers trust AI agents. These aren't fringe users. They're a substantial market segment with real spending power.

But trust varies by use case. Consumers are more comfortable with agents handling simple, low-risk tasks (product searches, price comparisons) than high-stakes decisions (major purchases, financial transactions).

Even more telling, 43% of consumers believe brands will eventually market directly to AI agents. The question isn't if, but when and how retailers will adapt.

‍

What strategies are working for companies that are building trust with consumers?

Retailers earning consumer trust are transparent about when and how they're using AI. They give customers control (clear preferences, easy opt-outs, visibility).

And they start small. Testing with simple recommendations, then price monitoring, then more complex decisions. Building trust incrementally through smaller autonomous actions that prove the system works.

Retailers facing skepticism are trying to do too much too fast, or can't visibly articulate their governance approach.

‍

How is consumer discovery and search behavior changing? What does this mean for retailers?

Discovery is fundamentally changing. Consumers, especially younger ones, are retreating from traditional social media to closed communities (WhatsApp groups, Discord servers, trusted creator communities). Word-of-mouth is winning over algorithmic feeds.

There's real concern about the sustainability of current social media models under the weight of AI-generated content and misinformation.

For retailers, this means diversifying your discovery strategy. Focus on building direct relationships with customers and reaching them through authentic creators in closed communities.

It also means preparing for a world where consumer agents do the discovery and purchasing. How will your products be found and evaluated by AI agents?

‍

What should retail leaders actually be doing right now?

Three priorities:

First, focus on achievable AI-powered automation before jumping to fully agentic. Customer service, inventory optimization, content generation deliver value today and build the foundation for more autonomous capabilities later.

Second, build your governance framework now. Clear decision-making parameters. Escalation protocols. Monitoring and quality controls. The credible players are treating governance as core strategy, not an afterthought.

Third, start experimenting in contained environments. Pick one area, test agentic capabilities with proper controls, learn what works, build confidence, then expand methodically.

You don't need to wait for full autonomy to capture value. Retailers can build competitive advantage today by implementing governed automation.

‍

What's the realistic timeline for mainstream adoption? When should retailers expect different capabilities?

For customer service and operational optimization: two to three years of test-and-learn with gradual autonomy increases. This will become table stakes for staying competitive in retail.

For complex agent coordination across your business (inventory agents working with pricing agents working with marketing agents): three to five years for broader adoption.

For true autonomous decision-making where agents make strategic business decisions independently: five to seven years minimum, and even then, significant human oversight required.

Consumer-facing agentic experiences are on a similar timeline. Shoppers having AI agents make autonomous purchases will take time because of trust factors.

‍

As we wrap up, what's your main advice for retail leaders navigating the hype around agentic AI?

Lead AI transformation with confidence, but confidence grounded in responsible implementation.

The opportunity is real. The companies that will win are those building strong foundations (clear governance, transparent practices, proper controls). They're not rushing toward full autonomy, but they're not paralyzed either.

Start with what works today. Customer service, inventory optimization, marketing automation deliver measurable value. Build trust through smaller autonomous decisions before tackling bigger ones.

This technology will reshape retail, but on a timeline of years, not months. The retail leaders who succeed will balance innovation with pragmatism, understanding that the real opportunity lies not in rushing toward autonomy, but in building retail systems that earn the trust required to scale responsibly.