6 takeaways from Chatbotconf 2017 in Vienna

At ChatBotConf 2017, more than 25 speakers presented new insights, features and best practices about chatbots in messenger apps. At the conference, which is organized annually by the messaging platform oratio in Vienna, developers, startups and companies exchange views on topics such as messenger marketing, conversational design and natural language processing in talks, panels, workshops and meetups.

Here are six takeaways from Facebook, Twitter and Chatfuel, among others:

1. Messenger marketing tops email marketing

Consumers spend a lot of time in messenger apps. Dmitriy Kachin, Head of Partnerships at Chatfuel, in the master class at ChatBotConf.

With chatbots, companies can use messenger apps as a marketing channel and benefit from the wide reach and direct communication with customers. In addition to being used in innovative marketing campaigns (such as Jäm Bot), chatbots are also successfully used in classic direct marketing and achieve better performance there than emails!

Dmitriy Kachin from the Chatfuel platform presented the average read rates and click rates of messages sent with chatbots via the platform at the conference. On average, messages from chatbots in Messenger are read by 85% of recipients, while promotional emails are usually only read by 20-30% of recipients! The click rates in messenger marketing are also convincing — more than 40% of users then click on further links in the message.

2. Use Webview

Using the webview for option lists (left). Helen Tsang, Engineering Manager at Facebook Messenger at CBC17 (right, photo: Jan Hubrý/chatbotconf)

Pure chat interaction is not suitable for all use cases in a conversation and can lead to a poor user experience. Helen Tsang from Facebook presented some best practices for the UX of chatbots in Facebook Messenger.

Among other things, she recommended that chatbot developers make it easier to query data using web views in order to avoid a “death by decision tree.” Using option lists in Webview, users can also make changes to their selections more easily. The webview should be integrated into the conversation like a widget and not be used as an independent web app.

3. Balance between fun and utility

Ashley D'Arcy from Poncho on the Main Stage on October 3, 2017 in Vienna.

The character of a chatbot or voice assistant must match the brand or the product and its environment — a topic that came up repeatedly at the conference. Ashley D'Arcy develops the personality behind the Poncho chatbot, a cat that sends users creative weather forecasts on Facebook Messenger every day. She presented how the team behind Poncho uses A/B tests to check which tonality, media and elements create the desired balance between the use of the bot and its entertainment factor for the user.

In Facebook Messenger, users usually chat with friends and family, which is why a chatbot in the messenger app can certainly show personality if it fits the brand or purpose of the bot.

Helen Zeng from Slack also addressed the issue of inappropriate and untargeted tonality in her talk at the conference. Slack is a messenger app for companies, which is why chatbots on the platform are aimed at business users.

Irrelevant or long texts and entertainment elements such as jokes or GIFs can be perceived as inappropriate or annoying and result in the bot no longer being used.

4. Customer Service, Transactions, and Tools

Michael Finkler from API.ai, a platform for natural language processing from Google, spoke on the main stage about conversational experiences and sees the successful use of chatbots primarily in customer service, for transactions and as useful tools in everyday life.

Potential savings through automating dialogs with the help of chatbots in service have been an issue in the industry for some time. Finkler also sees the use of chatbots in e-commerce as a personal ordering channel or in the smart home sector to control devices.

5. Twitter relies on chatbots

“Convert conversations into relationships.” Joe Rice from Twitter presented chatbots from well-known US customers, such as Wendy's, Denny's, Starbucks and United.

Until now, bots on Twitter have mostly been mentioned in negative contexts — this is about to change. On the morning of the conference, Joe Rice presented the use of chatbots for companies to automatically answer customer inquiries via direct messages on Twitter. If necessary, an employee can take over the conversations at any time.

The Twitter chatbots presented by American fast food chains such as Wendy's, Denny's, Starbucks or the airline United Airlines are already very well received by users and, in addition to customer service functionalities, also offer the opportunity to place orders, find branches in the area or take part in surveys and quizzes.

Users can access additional functions via a menu within the message window, similar to bots in Facebook Messenger. Companies can link to their chatbot using call-to-action buttons under tweets and thus promote it. In addition, users can access profiles via a message button and buttons on external websites to chatbots. Although Twitter is only actively used by around one million people in Germany, the potential of chatbots is enormous, as more than 80% of global users send inquiries and complaints to companies via the platform.

6. Added value for users and companies

In a fintech & insurtech panel in the afternoon with representatives from startups, banks and insurance companies, initial approaches and business models of chatbots in the banking and insurance industry were discussed. A key takeaway of the discussion round was that existing products should not be transferred to a chatbot and that new products with added value were needed specifically in the messenger channel.

On November 6, another panel on chatbots in the fintech sector will take place as part of FinTechWeek Hamburg. You can sign up for free here.

The development of conversational interfaces is still in its infancy, but the large technology companies in particular are significantly expanding the sector. Unfortunately, only a few use cases were presented at this year's ChatBotConf, which focus on added value for users and companies. The use of chatbots within companies and in the B2B environment has also received little attention, although this area offers great potential.

The still young market is growing rapidly — many brands and retailers are already gaining initial experience with conversational interfaces and optimizing their own chatbots and voice assistants. It is therefore quite possible that ChatBotConf 2018 will have a stronger focus on practical use cases and the topic of user acquisition.

‍