Written by Greg Ahern, Founder and President of Ometrics
Chatbots have become common enough on websites that people are not confused or intimidated by them anymore. Users even look for the chatbot icon and know exactly what it means. People expect efficiency in their day-to-day business and this is one way to make sure your website provides it.
A well-designed chatbot will engage your site visitors and allow deeper interaction with prospects and help customers support issues. Here are three benefits:
Increase ecommerce sales, units sold, and average order value
Create qualified business leads and appointments
Reduce support tickets so your support departments can be more productive.
AI chatbots collect a lot of data that provides marketing insights about what your customers are looking for, competition, site bugs, coupons issues, brand issues, etc. Every conversation is retained and analyzed.
Chatbots connect to many systems. User inquiries can be sent to sales CRMs and support ticketing systems. Ecommerce sites can have their data read by the chatbot and displayed in the chat window along with order status and other information. Below is an example of an ecommerce chatbot.
Types of Chatbots
1. AI Chatbots
An AI chatbot has the ability to understand what is being asked and reply in a helpful way. The most common way to interact is via typing but there are voice and SMS interfaces as well. AI Chatbots are relatively easy to create and teach. Chatbots do need to be maintained. AI works by having the chatbot continue to learn. “Human in the loop” compared to “machine learning” assures the chatbot is not learning the wrong way to respond from a social perspective. They also don’t automatically understand all the nuances of language, misspellings, or the way a texting culture changes the language “(U R” instead of “you are”). The AI chatbot has to be taught all of these.
2. eCommerce Chatbots
eCommerce chatbots are great for helping out with sales conversions for complex products or sites with a lot of SKUs. The user may not want to browse your entire online store but instead has a good idea of what he or she wants, and needs help finding it. The chatbot can help the user, from shopping assistance to adding to the cart, and with coupons and answers to support questions. If you want, you can still use live chat as a backup when the chatbot doesn’t know the answer or when the shopper can’t decide and needs another opinion.
A chatbot can remember you from that last time you visited and pick up the conversation where you left off. It needs to be able to answer questions about products, order status, returns, sales, coupons, and shipping. The more categories of products you offer, the more complex the programming is to direct customers to the type or category of products they may want. The number of products is not as much of an issue since all the products are loaded into the AI at once from the ecommerce platforms like Shopify and others.
3. Support Chatbots
A lot of support questions are very simple to answer and don’t need a live person to answer. Support chatbots don’t need to sleep (even while being programmed with new intents) so they are on duty around the clock. This keeps support staff focused on higher level tickets. For many applications, having live chat as a backup is an option to help when the user has inquiries that the chatbot doesn’t know how to handle. This combination is still less expensive than having to hire enough staff to handle lower tickets. The more options there are for troubleshooting, the more complicated the chatbot can be to build. Its complexity is also dependent on how much you want the chatbot to solve before turning the user over to a live support agent.
4. Lead Chatbots
LeadBots engage visitors at a higher rate and make it easier to gather information than online forms. They not only have qualifying questions but can also set appointments. The conversational experience makes it more welcoming for the user and more likely the visitor will answer the questions and give qualified information.
LeadBots are prebuilt and can be launched in seconds. They can also be customized for a particular business need. As with all bots the data can be sent directly to sales teams and any CRM. LeadBots are also on social sites like Facebook and other platforms.
5. Live Chat
Live chat is great for complicated inquiries that are easier for a human to figure out and or answer. Examples might be confidential medical questions, shopping for a picky gift recipient, or complex applications of a product or service. The best application is to have the AI Chatbot engage first with the user then ask the user if they would like to talk to a human. Web visitors prefer AI Chatbots over Live chat by 63% this is often because they feel the conversation is more private and less salesy. Live chat is often confused with an AI chatbot so it's important that your chatbot identifies itself as a chatbot, not a human.
Get Started on Your Chatbot
Building a chatbot depends on the type, technology, and complexity of the application. Simple chatbots can be built in minutes. The average time to build a complex chatbot ranges from 3-4 days to a couple weeks. Once your bot is built it can take at least a couple months to become smarter depending on your site traffic and user engagement.
If you are building a chatbot from scratch costs can range from $20k to over $100k. If you are using tools like Ochatbot the cost can be a couple from $35 a month to $495 a month.
The first step to building a chatbot is to step back and think about the goal for the bot and how you want to measure its success. Then imagine talking to a customer and about the different types of conversations and branches of those conversions you could have. From this you can start building out your first chatbot. The narrower the intelligence of the bot, the smarter and simpler it can be to create and train.
Creating a decision tree type diagram is the easiest way to begin building these types of chatbots. The decision tree spreadsheet then becomes your big picture and the map for you to write the intents of the chatbot.
Start with a simple bot then grow the bot’s intelligence over time. Spending days in meetings about what you think your users will ask is often not as accurate or efficient as just letting your visitors ask, and then training the bot daily. This allows you to launch quickly and can save a significant amount of time and development money.
About the author, Greg Ahern
Greg Ahern, Founder and President of Ometrics®, The Marketing Alliance Chatbot Partner, is a fanatic about conversion rate optimization, AI chatbots, and lead generation. Greg has been a successful Internet entrepreneur since 1994. He speaks at conferences and webinars and has built a number of internet businesses, including web marketing, web development, and internet lead generation, which have been successfully acquired. Greg was a Denver Chapter Leader for the Digital Analytics Association. You can follow Greg on Twitter @gregahern and join his CRO GrowthHacks Group on Slack.
