- 4comments
- 142comments
- 23comments
- 207comments
- 42comments
- 174comments
- 3comments
- 27comments
- 59comments
- 428comments
- —discuss
- 95comments
- 42comments
- 35comments
- 54comments
- 159comments
- 639comments
- 3comments
- 265comments
- 101comments
- 177comments
- 87comments
- 11comments
- —discuss
- 240comments
- 56comments
- 163comments
- 168comments
- —discuss
- 25comments
Problem: Live chat has the highest satisfaction rate over email and phone for customer support [1] and is the preferable channel. As companies like Facebook, Telegram, etc make it super simple for customers to connect directly with businesses, having someone in front of a computer to answer questions is not a scalable solution. Ecommerce owners should not spend their time/money building and training a bot.
Solution: We will provide ecommerce owners a platform where they can get their own support bot up and running with a few clicks. Our service will connect to their ecommerce solution, gather the necessary data, learn their FAQs and with all these information the bot will try to answer customers questions. As more ecommerce companies join the platform and more questions are asked, the bots will become smarter.
Who are we? I am Guillermo, a Software Engineer with over 10 yrs of experience working for big companies and started a handful of startups. I have a successful bot running on Slack and the lessons learned there will help me build this platform. Angeli, my wife, has been a PM in the software industry for over 5 years. She has been running an ecommerce for about 4 years with annual revenue in the 6 digits. We recently quite our corporate jobs and became digital nomads, ready to work on this full time.
[1] https://econsultancy.com/blog/63867-consumers-prefer-live-chat-for-customer-service-stats/
Will the chat bots self identify as chat bots to customers?
This is something that should probably be tested and see how people react. "chat bot" is something that is common for us but not for non-technical people and could actually hurt more than help. The experience for the customer should be quite smooth and once the bot is not able to answer a questions something like "Let me get back to you about this" could mean that the questions is being redirected to a real human. Again these are assumptions and data will give us the real answer.
This is a interesting field to go into and the potential market is large. Is it more efficient for a user to use a chat bot instead of searching a FAQ or other current alternatives?
Imagine that you just made a purchase at an online store and you want to know how long is it going to take for you to get the product. That information is available on the site, somewhere. You can do that or you can open your favourite messenger platform and ask "hey how long does it take to get an order", just like you ask your partner what to get from the supermarket. It is a way people are use to communicate. The result is the same but the customer satisfaction is higher. The bot will be able to answer that kind of questions quickly without human intervention. Is more efficient for the customer and the company.
I help out with the online chat for my wife’s business sometimes and it is amazing how many different types of questions you get asked. Sure around 50% are one click away in the FAQ, but around 50% are unique in some aspect.
The key to making this work will be the quality of the data behind the bot and the quality of the AI. These are hard problems, but if you can make them work then you will have a very impressive product.
Hi there, funny enough this idea came up helping my wife with her online store. Yes, it is a hard problem and we are up for the challenge. Initially will be a hybrid of manual answers and automatic while we train the algorithms But with enough questions and multiple stores we should be able to have a very good knowledge base.