The difference is what happens after the message.
These are common product patterns, not rigid labels. Some chatbots also use tools and retain context. Evaluate the actual journey, integrations, and controls rather than relying on what a vendor calls the product.
| Capability | Support-focused chatbot | Agentic shopping experience |
|---|---|---|
| Primary job | Answer questions or route support. | Help complete a shopping goal. |
| Context | Often session-limited or support-focused. | Retains product constraints and journey state. |
| Product results | May return links or text. | Renders grounded products, variants, comparisons, and controls. |
| Actions | Usually redirects or hands off. | Can perform merchant-approved cart and workflow actions. |
| Interface | Conversation window. | Conversation plus search, cards, filters, product pages, and cart. |
| Success metric | Containment or response time. | Discovery quality, decision progress, cart actions, and conversion. |
When a chatbot is still the right tool.
- Answering policy, order, delivery, return, or account questions.
- Routing a customer to a human or existing support workflow.
- Handling narrow, repeatable intents that do not require product reasoning.
- Providing service outside the shopping journey on messaging channels.
When an agentic shopping experience adds value.
- The catalog is difficult to navigate with keywords and static filters.
- Customers need help translating a goal into product attributes.
- Comparison, compatibility, fit, or tradeoffs matter before purchase.
- The merchant wants the experience to advance toward cart or checkout.
- The business can expose accurate data and controlled actions.
Questions to ask a vendor.
- 01
How are answers grounded in live catalog, inventory, price, and policies?
- 02
What happens when the request is ambiguous or unsupported?
- 03
Which actions are possible, and how are permissions and confirmations enforced?
- 04
Can the interface update product results and controls outside the chat window?
- 05
How is incremental commercial impact measured against a control?
- 06
What data leaves the merchant environment, and how is it retained?
Frequently asked questions.
Is every generative AI chatbot an agent?
No. Generating natural-language responses does not by itself provide context, tools, permissions, or reliable action execution.
Can one system handle support and shopping?
It can, but the data, workflows, success metrics, and risks differ. Teams should define each journey rather than treating every conversation as the same problem.
Sources and further reading.
External research and technical guidance provide context, not evidence of ScaleWave customer results or vendor partnerships.
- Features and capabilities of AI Commerce Search (opens in a new tab)Google Cloud Documentation
Product catalog, event, search, and customization capabilities.
- Conversational agent user experience guide (opens in a new tab)Google Cloud Documentation
Interaction guidance for voice, text, images, search, and product results.
- How Target is enhancing retail discovery with Spanner Graph (opens in a new tab)Google Cloud Blog
An August 2026 article by Target engineers on connected product data. Target is not presented as a ScaleWave customer.