How is a chatbot flow different from a decision tree?

A decision tree is a type of chatbot flow where every branch is explicitly authored based on exact customer inputs. Modern chatbot flows built on LLMs are more flexible — the model interprets natural language inputs and routes them to the right branch based on meaning, not exact matches.

How many chatbot flows should a Shopify store build?

Build flows for your top five to eight support scenarios by volume — those cover the majority of interactions. Trying to build flows for every edge case is counterproductive; the AI's general language capabilities handle long-tail scenarios well without dedicated flows.

What happens when a customer goes off-script in a chatbot flow?

A well-designed flow handles off-topic messages gracefully: it answers the diversion if it can, then returns to the active flow. If the customer clearly wants to abandon the flow, the system acknowledges it and resets to the main intent detection layer.

Glossary

Chatbot Flow

A chatbot flow is a structured, branching sequence of conversation steps — including prompts, conditional branches based on customer inputs, data collection points, and resolution actions — that guides an AI or scripted chatbot through a specific support scenario toward a defined outcome.

Book a demo See pricing

What it means

Key insight

A well-designed chatbot flow is like a best-practice support script that executes perfectly every time — without the variability of human execution.

Chatbot flows formalize what a skilled support agent does instinctively: understand the customer's need, collect any required information, check eligibility or conditions, and execute a resolution. In practice, a flow for 'process a return' might look like: (1) confirm the customer wants to return an item, (2) collect or verify the order number, (3) ask which item is being returned, (4) check return eligibility against your policy, (5) if eligible: confirm return method and generate return label; (6) if not eligible: explain why and offer alternatives. Flows vary in rigidity. Rule-based chatbots follow rigid flows where every branch must be pre-authored. LLM-based systems can follow flexible flows where the model handles language variation naturally within a defined structure. Hybrid approaches — LLM for language, explicit flow logic for process control — combine the best of both, ensuring structured processes execute reliably while keeping conversations feeling natural.

Why it matters

Chatbot flows encode your support best practices in a repeatable, scalable form. Every customer who contacts about a return gets the same well-structured experience as if your best agent handled it — because the flow embodies what your best agent would do. For Shopify merchants, the highest-value flows to build are for your highest-volume, most structured scenarios: order status lookup, return initiation, refund status check, and discount code application. Each well-built flow directly reduces handling time, increases first-contact resolution, and creates a consistent brand experience.

How Bookbag helps

Flow Template Library

Bookbag ships with ready-to-use flow templates for the most common Shopify support scenarios — returns, order lookups, shipping questions — that merchants can deploy immediately and customize over time.

LLM-Flexible Flow Steps

Within each flow step, Bookbag uses LLM understanding to handle natural language variation, so customers don't need to follow a script — they can answer naturally and the flow still progresses correctly.

Flow Performance Analytics

Bookbag tracks completion rates, drop-off points, and resolution outcomes for each flow — surfacing exactly where customers abandon and which steps generate confusion so you can iterate on the design.

Frequently Asked Questions

See Bookbag in action

Join the ecommerce teams resolving more tickets, answering 24/7, and turning support into a revenue channel with Bookbag.