We came across the Epytom chatbot, it’s implemented on both Facebook Messenger and Telegram. The team at 1080bots is studying ecommerce chatbots to improve our delivery of productive ecommerce conversational experiences. The Epytom chatbot is not a store, it does not sell, but it helps you buy if you follow the recommended flow.
The Epytom site is all about creating a limited wardrobe that begins with the items you already own. The chatbot begins with that introduction and without prompting gives you an inspiring outfit for the day. Great initial interaction and good use of images, buttons, and emoticons.
The Epytom chatbot then provides a set of easy options to bring up more style tips, to build a weekly tip sheet or to personalize the service. Nice and simple, well guided and not at all intrusive.
The chatbot asks for a location and quickly brings up a personal checklist of style essentials as a well laid out chart that can be downloaded and printed. So far the engagement was great, very low effort on the user’s part and high value in terms of the result.
The chatbot is now making good recommendations, however, on the website, the value proposition seems to continue to help us get “true savings”.
We looked for somewhere on the website where we might find out how to buy the recommended items but no luck. We then looked on Facebook, nothing there either.
Finally, we went back to the Epytom chatbot in Facebook Messenger. Early on in the dialog, alongside the picture of Steve McQueen in the movie The Great Escape, we saw a ‘Get It’ badge. We went to the menu, but there was nothing there. We tried typing in that we were ‘Looking for a Jacket’ – no luck and it wasn’t able to handle the unstructured text. Our final try was to see if it could help with ‘Chelsea boots’ but no luck.
We enjoyed the simplicity and concept behind the Epytom chatbot, However, whereas we normally complain about a chatbot getting you to a product page too fast, this chatbot left us more than a little frustrated by not allowing us to make a purchase, or even recommending retailers. This might be part of the future strategy for Epytom, so we’re not going to judge just yet.
If you’re thinking of creating an ecommerce bot that deeply integrates into the back-end retail systems and can handle natural language requirements then drop us a line at email@example.com.