Implement smart question and answer intelligence in chatbots
1. Collect all information required for implementing chatbots on your website, like the service you'll use and how you'll harvest email addresses from it.
Select a chatbot service, like Facebook Messenger, Drift, services such as collect.chat, WSChat, Botsify, Chatfuel, FlowXO, Landbot.io or IBM Watson Assistant. If you have access to developer resources, AWS has a Chatbot that you can use to create smart bots for your website. If you use WordPress, install a chatbot plugin for the service that lets you select custom URLs for deployment. Create a new spreadsheet titled Intelligent Chatbot Implementation with two worksheets and name them Display URLs and Questions & Answers. Select a service for saving email addresses once users enter them in the chatbot. Check the documentation to ensure that it can be plugged into the chatbot service you’ve chosen. Create a test page on your website to implement the chatbot before going live.
2. Design a conversation flow for the first chatbot that includes up to three levels of questions and follow-up questions.
Open the conversation flow designer for your chatbot tool. If the service you selected does not provide one, you can use Google slides or Figma. Create a flow chart with up to 3 levels and all possible answers for each of these levels. For the offer to connect with a rep, ask permission to collect user details such as their name and email address. Follow up with the questions, and connect the inputs with the email / CRM service selected in step 1 Repeat as necessary for each page and for each step in the node. Set up the goodbye message and save the chatbot conversation flow.
3. Customize the branding for the chatbot, like the chatbot icon, colors, and name.
Choose your brand colors for the display of the chatbot. Many tools will request RGB color codes for the branding. If you don’t know these codes for your brand, use a Chrome extension called Eye-Dropper to find them. Install the extension, open your website in the browser, and click on the extension tooltip anywhere on your page, and it will reveal the hex code of the color you clicked on. Choose an icon to use as the chatbot’s identity. Options include the square version of your logo, pictures of your customer support team, or the generic bot images that are available with most chatbot tools. Select the initial loading size and location. If the option is available, give your chatbot a friendly name.
4. Test the chatbot implementation on your test page.
Submit the chatbot to go live in the service you’re using. If you’re using WordPress, configure the chatbot to run only on your test page. If you’re using another service, paste your chatbot’s code into the test page. Open test page. Test the chatbot by answering its questions and validate its responses against the designed conversation flow.
5. Deploy the chatbot on the live environment of your website.
Paste the pop-up code in each URL of the page. If you use WordPress, paste the code into the header include file or ask a developer to implement it for you. Test it on a few different website pages to make sure it is working properly.