The AI Chatbot is an AI-powered chatbot widget that marketers add to their websites. It provides real-time answers to end-customer queries using connected knowledge sources and can create support tickets in Maropost Service Cloud (MSC) when an issue cannot be resolved in chat.
The AI chatbot is designed to reduce repetitive support workload, improve response times, and ensure unresolved issues are logged as tickets with conversation context.
Key Terms
Term | Definition |
|---|
Marketer | An individual on a business's behalf who uses MSC and embeds the chatbot on their website. |
Customer | A website visitor who interacts with the marketer's chatbot. |
Ticket | A support request created in MSC from a chatbot conversation. |
MSC | The Maropost Service Cloud platform, where marketers use the ticket management feature. |
Prerequisites
- You have access to MSC with permission to create and manage chatbots.
- You have a website/store URL where the chatbot widget will be embedded.
- You have branding assets ready (for example, a logo and a primary brand color).
This feature is exclusively useful for:
- Marketers/support leaders who want to reduce repetitive tickets and improve response times.
- Support admins who configure branding, knowledge sources, prompts, and business hours.
- Support agents who need predictable ticket creation behavior and a clean context on tickets created by the chatbot.
Feature Breakdown
The AI chatbot's feature breakdown is as follows:
Chatbot Management in MSC
The Chatbots area provides a list view for managing chatbots connected to stores. This includes:
- A list of chatbots with columns for store, store URL, status (with an activate/deactivate toggle), conversations, created on, last modified, and actions.
- An Actions menu per row with the following options:
Edit: Open the chatbot configuration.
Archive: Remove the chatbot from the active list (see Archiving and Managing Chatbots).
- A Create New Chatbot button in the top-right corner to add a new chatbot.
- An empty state screen displaying “No chatbots yet” with a Create New Chatbot button when no chatbots exist.
- An Archived Chatbots view that lists archived items with their store, store URL, status, conversations, archived on date, and a restore action per row.
Customizing Your Widget
You can configure your chatbots by clicking on the Create New Chatbot button. The Customize Your Widget screen includes a left navigation and a live Preview panel. Configuration areas in the left navigation include:
- General
- Appearance
- Knowledge Base
Knowledge Base Sources Used for AI Responses
The chatbot answers questions using enabled sources in the knowledge base. Available sources include:
- Frequently Asked Questions
- Help Articles
- Company Website URLs (Pages)
- Support Docs
The Add New Source area supports:
- Website URL: Add a page URL to the index. The Maropost user provides their website's homepage URL, from which we automatically crawl and identify all relevant pages. They can then select the desired pages for text extraction, as well as manually input any additional pages that may not have been detected.
- Upload Files: Upload a file (max 5 MB; supported formats: PDF, TXT, Doc).
- Knowledge Base Questionnaire: Fill out a guided questionnaire to generate a knowledge base automatically. The questionnaire covers your company description, products, target audience, and brand tone; your shipping and order processing policy; how you handle delivery problems; your policy for modifying or cancelling orders; and your full returns and exchanges policy.
Once a source is uploaded, the system displays its processing status (e.g., "Processing and indexing content… This may take a few moments."). Each connected source shows the item count and last-synced date and can be individually enabled or disabled via a toggle or deleted.
Note: A source that is disabled will show a tooltip: "This source is inactive, and the chatbot will not use this information when responding to users." A source that is indexing will show the following: "The AI is currently processing and learning from this source to build its knowledge base."
Customer States and Ticket Creation Behavior
The chatbot determines the customer’s state and applies ticket creation rules accordingly. Do remember, for Iteration 1, manual agent handoff is not included. For complex queries, the chatbot creates a ticket so an agent can follow up in MSC.
User State | Chatbot Entry and Greeting | Order Information Access | Ticket Creation Behaviour | Data Stored |
|---|
Anonymous user | Generic greeting | Not available | A ticket can be created if unresolved; explicit consent is required | Anonymous session ID and chat transcript |
Guest checkout-only user | Generic greeting | Not available | A ticket can be created if unresolved; explicit consent is required | Anonymous session ID and chat transcript |
Unauthenticated account user | Generic greeting | Not available until login | A ticket can be created if unresolved; explicit consent is required | Anonymous session ID and chat transcript |
Authenticated account user | Personalised greeting if context is available | Available | A ticket is automatically created for unresolved issues | Anonymous session ID and chat transcript |
Chat Widget: Sample Conversation Flow
The chatbot widget displays in the bottom corner of the website. A sample interaction flow for a guest user includes:
- The widget opens with the greeting "How can I help you today?" along with configured Quick Prompt buttons (e.g., Track my order, Email invoice, Ask about products).
- The customer selects a prompt or types a message.
- The chatbot responds with follow-up options (e.g., "Sure, what is your question?" with sub-options: “Where is my order?” / “My order hasn't arrived yet” / “I want to change my delivery address”).
- The customer selects an option (e.g., "My order hasn't arrived yet"), and the chatbot provides the expected delivery date and any relevant details.
- For requests such as Return Item, the chatbot can display an image attachment (e.g., a product photo) alongside the response.
Create a Chatbot
- To access this feature, head over to the navigation panel and click on Service > Chatbots.
- Click on the Create New Chatbot button. If you haven’t designed a chatbot previously, click on the Create Your First Chatbot button instead.
Configure General Settings
- Open the chatbot you want to configure.
- Navigate to Customize your widget.
- Select General from the left navigation.
- Complete the fields under General and save the changes post configuration. These include Store Name, Address 1, Address 2, Instagram link, Store URL, and a Store Type dropdown (e.g., Neto or Other). For Neto stores, enter your API key in the field provided and click Connect to link your store. A set of team member avatars will appear to confirm the connection.
Configure Appearance Settings
- Select Appearance from the left navigation.
- Under Brand Logo, click Upload Logo and upload your logo file. The recommended image size is 160×160 px, PNG or SVG.
- Under Primary color, select a color swatch from the preset options (black, blue, purple, teal, green, yellow, orange, pink).
- If required, enable Use Custom Color and set a color using the color picker (a default Vuetify 3 picker). Use the gradient canvas to pick the hue and brightness and the hue and opacity sliders to set the base color and transparency. You can also enter a hex value directly in the #RRGGBB field or adjust the R, G, B, and A values individually. Previously used colors are shown in the saved swatches panel below.
- Under Widget configuration, choose the widget position either on the Bottom left or Bottom right.
- Set the welcome message that appears when the widget loads (e.g., "Hello! How can I help you today?") and save the changes made.
Configure Knowledge Base Sources
- Select Knowledge base from the left navigation.
- Review the Connected Sources list (item count shown top-right).
- Enable or disable each available source using its toggle: Frequently Asked Questions, Help Articles, Company Website URLs (Pages), and Support Docs.
- Under Add New Source, add a knowledge source by clicking Website URL to enter a page URL to crawl and index, Upload Files to upload a file (max 5 MB; supported formats: PDF, TXT, and Doc), or Knowledge Base Questionnaire to open the guided questionnaire form and auto-generate content.
- Wait for the system to finish processing and indexing the uploaded content, and save the changes.
Archiving and Managing Chatbots
- To access this feature, head over to the navigation panel and click on Service > Chatbots.
- Open the Actions menu (⋮) for the chatbot row you want to archive. Click Archive.
- In the Archive chatbot dialog box, review the warning message that states the chatbot will be disabled immediately and removed from the active list; conversations will remain accessible for 30 days, after which the chatbot will be permanently deleted. Click Archive to confirm or Cancel to go back.
- To view archived chatbots, choose the Archived option from the dropdown list. Review the archived list; i.e., each row shows store, store URL, status, conversations, and archived on date details. To restore an archived chatbot, click Restore under the Actions column of the relevant row. The chatbot gets activated after restoring it.
Connecting AI Chatbot with Neto
You can utilize the Maropost add-ons available in Neto to link your Service Cloud (which includes the AI chatbot interface) and Neto accounts.
- Assuming you’ve already installed the Maropost add-on on your Neto account, log in to your Neto Control Panel, head over to Addons > Maropost, and install the Service Cloud add-on that Maropost has to offer.
- In the pop-up window that appears, enter your Maropost Account ID, then click Install. This account ID is required to enable all Service Cloud features, including the AI Chatbot feature. You can access this account ID under the Company Information tab of the Account & Billing feature. The account ID is clearly visible under the Account Information section.
- Click on the Configure button to link the Service Cloud add-on to your Neto account.
- You will then be prompted to sign in using Maropost Identity or redirected to link an existing account. After this step, the Service Cloud add-on gets installed on your Neto account.
- Once installed, proceed to create the chatbot from your Maropost account via Service > Chatbots.
