Chatbots were the hottest thing for a while there. In 2017, they dominated much of the public discussion surrounding technology, having an important presence in events like Mobile World Congress. However, after that initial breakthrough into public consciousness, excitement has somewhat died, with some even claiming that chatbots are dead.

The truth is far from that. While the market certainly experienced some setbacks and disappointments regarding chatbots, there’s no need to be such a fatalist about it.

First, some context. Months after chatbots made a splash, the future they had promised seemed to be farther away than previously thought. Digit, an automated financial savings service, was a company that experienced that reality firsthand. “Chatbots are dead,” Digit founder Ethan Bloch declared an interview to Inc.

Originally, the company had tried to go big on chatbot implementation. Instead of using a conventional interface, users texted with a bot that asked questions to set their preferences for the service. After a while, though, it became apparent that users preferred a traditional experience with menus and finger taps. Digit then decided to revert to that format.

Helpshift, a company that actually bases all of its business in chatbots, is also no stranger to the fall in popularity the technology has apparently suffered. Chief Strategy Officer Abinash Tripathy explained that most bots that failed did so because they relied “entirely” on natural language processing. In contrast, their bots are based on decision trees, where the conversation can branch out but is still somewhat pre-determined.

The greatest evidence of chatbots’ disappointing year, however, is the fact that almost no predictions about them came true. But that’s when we have to ask ourselves if they were truly predictions at all, or instead promises made by companies eager to sell their product and/or earn funding.

“I’m not even sure if we can say ‘chatbots are dead,’ because I don’t even know if they were ever alive,” declared Bloch. And he’s on to something, even if he doesn’t realize it. Because that’s the main issue: chatbots have just barely started to get traction, and expecting revolutionary results in a year or even two is just being quite unrealistic. We are simply not there yet. However, that doesn’t mean we won’t get there.

A popular chart by information technology firm Gartner that intends to explain the adoption of specific technologies has since become a standard for the industry, but one that is often ignored to give way to more exciting news. It’s called the Hype Cycle, and it perfectly explains what’s happening to chatbots right now.

The theory is simple: after the introduction of a new technology, there’s a ‘peak of inflated expectations.’ Inevitably, those unrealistic expectations then come crashing down to a ‘trough of disillusionment’. Eventually, advancements are made and the tech starts to evolve, culminating on a ‘plateau of productivity.’ Right now, it’s easy to see where chatbots stand.

Making a true conversational bot is no easy task, as it’s something that even giants like Google and Apple still struggle with. Likewise, expecting bots to have a strong user base like apps (“that chatbot everyone uses”) is assuming way too much. Instead, their future might lie in sectors like enterprise, customer service or marketing.

According to Gartner (the same company that came up with The Hype Cycle concept), chatbots are on the rise, and there’s no stopping them. By 2020, the company says that 25% of customer service operations will rely on virtual assistants. That same year, “over 50% of medium to large enterprises will have deployed product chatbots,” according to Gartner research vice president Van Baker.

In essence: chatbots haven’t died. We’re just grounding our expectations, and keeping them that way will be much better for the industry in the long run.