There is no Chatbot Bubble, Neo!
A Reaction to „Bursting The Chatbot Bubble“ on TechCrunch
This article is a reaction to „Bursting The Chatbot Bubble“ by Arun Uday on TechCrunch. I felt the need to share my thoughts about it with a community who I think is very eager to put something of value out into the world.
Article Summary
A short overview of the article, if you haven’t read it:
Arun writes how chatbots probably won’t live up to their promise. He thinks, the interface is too complicated in comparison to app UIs.
„Why would anybody want to replace an existing setup that is 100 percent accurate and takes less effort to use (i.e. UI-driven menus) with something that is inaccurate and requires more effort to use (i.e. chatbots)?“
He also mentions the danger of uncontrolled bot spam and finds it bemusing that many developers and businesses are jumping onto the train without evidence of a real user need for chatbots.
„That’s the reason I find it bemusing that not only are chatbot developers springing up like mushrooms in the monsoon season, but businesses are also actually eager to avail of their services sans any credible evidence of a real user need for them.“
At the end of the article, he puts his critique in perspective, saying that chatbots won’t inevitably fail:
„As with most advances in technology, once the current euphoria on bots dies down, we will see the emergence of real solutions to real problems coming out of the chat ecosystem.“
In the comment section, I also found a comment of Arun, stating that there probably won’t come anything good from opening messaging platforms for all developers:
„The central point of my article above though is that by making the chat platform completely open to any/every developer to do as they wish, we will see a proliferation of zombie bots which hardly provide any real end user value.“
Evolution = Trial & Error. My thoughts.
In German, we have this phrase „Schau ma mal dann seh ma schon“ which, roughly translated, means „Let’s wait and we’ll see“. It makes a bit fun of those who just sit and wait instead of deciding to act. I think if everybody in tech history would have said „I don’t have proof that there’s a user need for my idea — I better not pursue it“ and “Let’s see first if others will succeed with this idea.”, we would have few true innovations.
The chatbot industry is young, we’re all excited and dream of reaching all the users on all the chat platforms — that’s probably what Arun means with his critique on the many developers and businesses in the chatbot industry. And yes, maybe there are some people trying to make money with the high expectations, I certainly won’t deny that.
But in this phase I think we absolutely have to try every idea for a new bot to learn what works and what doesn’t. There are so many questions to answer:
- Will people use chatbots alone or in conversations with others?
- Do they like to solve problems conversation-based, instead of using an app or a website? If yes, which problems are more likely to be solved that way?
- Which new use-cases can we solve with bots?
- Which UI elements and tools for messengers do we need to make useful bots?
- How should a bot talk to a user? (More like a human or should he state clearly that he has AI / is just programmed for few functionalities?)
- How do we advertise bots?
- Should we develop bots cross-platform by default?
- Is the cognitive effort of the number of letters typed to complete a task via a chatbot really to be seen equal with the effort for an informed tap on a GUI? (As compared in the linked article by Dan Grover)
- …
We don’t know yet. That’s why there are so many people trying to make sense of chatbots — everybody has their own idea how they may work (and can be monetized), currently.
Yes, by trying to learn all that we will inevitably fail to create lasting value with many of our bots. But that’s not inherently bad, as we can then tackle the reasons for our failing and will know what to do with increasing certainty. Some day, we will have a solid understanding of why, what, how and when to do things with bots.
Let’s not forget that apple.com once looked like this:
UX & Design for websites once was in the same state as bots are now. The web was open to everybody to try and create something with or without real value. See how far we’ve come on the web! We know how to research and have learned a lot by trial & error. Now, we can provide many services vital to peoples everyday lives and businesses.
Imagine for example, if everybody had said „Why would anybody replace the convenient format of a newspaper with a website that requires more effort to use?“
True: Back then, a website wasn’t as convenient as the paper or magazine I could read at the kitchen table. But now, we have smartphones and tablets and we were able to enhance the news-reading experience with many things like sharing, commenting or displaying related articles,.. Society evolved and adapted to many ways of consuming news nowadays.
I think „Bursting the Chatbot Bubble“ with our doubts is the last thing we need to keep the industry innovating and creating better bot experiences. Yes, we have to discuss which use cases are relevant and what does and does not work with bots. Possible bot spam is certainly an issue we need to tackle. But we shouldn’t assume that these problems will prevent the bot industry from being successful. We can get those problems out of the way if we keep on learning and adapting our bot experiences. (e.g.: We can try to solve the spam-problem by establishing an obligatory „unsubscribe“-pattern for our bots, like in newsletters nowadays.)
Facebook & Co investing heavily in opening their messaging platforms for bot developers supports this learning process.
If you, Arun, made experiences with your chat app that could help the chatbot community grow, please share them. I’m sure they’re useful to everyone starting out in the industry. F.ex. you wrote:
„In fact, all experience suggests quite the opposite. When a customer has an issue, they would most likely want to skip automated response options and speak to a real person who can fix their problem — not chat with a semi-intelligent bot dishing out some canned responses.“
You have experienced a business scenario where 90% of users wanted to skip automated response options? We should really try to take this into account for similar situations. What business was that? Which questions did the users have? What age were these customers? How critical was it for the user to solve that problem (f.ex. to not lose their jobs / money)? … There’s always more to the story than „people just don’t like XY”.
I think sharing experiences and keeping on developing and learning is the best way to create more bots with real user value in the future.
It’s kind of a chicken-and-egg situation, isn’t it? We can’t get to the full potential of chatbots without beginning to create them. So let us put in the work and not think “Schau ma mal dann seh ma schon.” (“Let’s wait and we’ll see.”)
