Together with my community “ai-zurich”, I organize the AI for Business conference in Zurich once a year.
On our website ai-zurich.ch we inform website visitors throughout the year about the program, the invited speakers and other updates. A button is placed on each page that links to the ticket purchase landing page.
But when the users are finally on the ticket purchase page, in many cases they change their mind and don’t order a ticket or decide to wait and see…
With a chatbot on the ticket page, I was able to double the number of conversions, i.e. the number of ticket purchases, in a short space of time.
The small chat window only opens when a user has been on the website for more than two seconds. The reason is that the user should first have time to buy a ticket without the help of the chatbot. Only if he does not do this does the chat window open and the chatbot responds.
The chatbot asks the website visitor whether they already know the conference. If the user says no, the bot shows a video of the last conference.
As the conversation between the chatbot and the user progresses, the bot provides more and more information about the conference and regularly asks the user whether they have now decided to order a ticket.
The chatbot deliberately only works with a quick reply button and actually guides the user through a predefined conversation tree.
It is therefore a rule-based chatbot.
The quick reply button prevents the user from asking questions that the bot does not understand.
On the contrary, the conversation between chatbot and user can be easily predefined by me and steered in the right direction, so to speak. In this case, the right path means that every conversation ends, if possible, with the user clicking on the “Order ticket” button at the end.
The target group of the chatbot is basically the same as the target group of the ai-zurich website. These are managers and other business people who are interested in real-life applications and tips for using artificial intelligence in business processes or other people who want to learn more about the use of artificial intelligence. It is therefore assumed that the target group is basically technology-savvy. An analysis of Google Analytics figures has shown that website visitors are generally between the ages of 26 and 45.
A little tip, if you look at the demographic data of your website visitors in Google Analytics, you will get very good reports about the users of your website.
The image of our ai-zurich community is innovative, modern and friendly. The focus here is primarily on the community concept and the broad knowledge of content relating to applications of artificial intelligence.
Both the target group and the image of the community were taken into account when choosing the chatbot personality.
Both have led to the chatbot communicating in a more friendly, relaxed and young manner and addressing its users in the personal “you” form.
The chatbot was implemented with the chatbot tool aiaibot.com and only a few hours passed between dialog design and going online.
Once the conversion-oriented concept for the chatbot had been created, the dialogs had to be written, individual content elements such as videos had to be prepared and then the content could be compiled in the chatbot tool.
Some videos were already available. Overall, the dialog is rather short and concise. This is often the case with a chatbot as strongly conversion-oriented as this one.
Before the chatbot was made available to the general public, it was tested in a small group of 20 potential ai-zurich customers. A few adjustments were then made to the dialog and the chatbot was ready to go online.
As the chatbot is only supposed to work when a user is hesitating whether to buy a ticket or not, there were no major additional communication activities at the launch. Only two small LinkedIn posts were posted to publicize the chatbot and then the chatbot was supposed to try its luck itself, so to speak.
Analyses have shown that the bot is used by every third visitor to the ticket landing page. Please note that this chatbot was only published on the conference ticket page. And one in four users who tested the bot even ordered a ticket in the end.
The chatbot from my AI for Business conference is a successful example of a chatbot to support conversions. Alternatively, a pop-up window could have opened instead of the chatbot. However, pop-ups are rather unpopular with most users. A chatbot, on the other hand, that engages users in a conversation with the right personality and thus motivates them to make a purchase is much more successful in many cases, as this example shows.
If you have any further questions about this or other chatbots, please do not hesitate to contact me.
