We were promised chatbots — again, and again. These benevolent bots were supposed to automate and revolutionize customer service, paving the way to a new era of consumer engagement.

So what happened?

Some chatbots have certainly thrived — we’ve seen how e-commerce businesses such as Sephora, Pizza Hut, and Burberry have used them to connect their customers with products and drive purchases. Overall, however, the technology has not transformed the way most fundraisers on the non-profit side engage their fans and donors. Many chatbot cheerleaders have been left wondering what happened to the promise of tireless automated employees who could make it easier than ever for organizations to interact with their supporters.

“I’m sorry, Dave. I’m afraid I can’t handle all of your customer service needs yet.”

While reality has so far fallen short of the vision, the dream of chatbots is alive and well. Keep the following in mind as you decide whether a chatbot is right for your organization:

Chatbots have strengths and weaknesses — just like any other channel

Chatbots can be an effective and low-maintenance tool for non-profits. Our clients at the Sierra Club built a bot internally last year which allows supporters to learn more about and donate to various environmental and social causes — they’ve seen email-beating response rates (albeit with a smaller audience, of course), some great friend-to-friend referrals, and nearly 100% action-to-open rates.

And if you spend 15 minutes playing around with Chatfuel (a platform for developing Messenger bots), you’ll see that creating the interface is easy. What’s holding chatbots back is the state of artificial intelligence (AI), whose job it is to interpret what the bot should do with all that information it’s collecting.

AI will catch up, but in the meantime, I suggest re-evaluating where chatbots work best. It may be painful to accept that broad, open-ended journeys are not suited well to chatbots just yet, but chatbots can be effective when paired with well-guided journeys through largely-static option trees (often called Conversational Interfaces, or CI).

However, the prevalence of voice assistants like Alexa is shaping how people expect to interact with bots. Our clients at Idealist.org have found that their bot, which uses a CI design, can be frustrating to users who expect it to be able to respond to anything they may type. Advances in natural language processing (NLP) — teaching machines to understand more conversational inputs from humans — are promising, but implementing NLP can be time- and resource-intensive for organizations.

With many orgs hesitant to invest heavily in bots, many of today’s Facebook Messenger bots lean on buttons and prompts versus freeform text entry. As much as marketers want omniscient chatbot agents that can handle any input seamlessly, users just want to get to the end state as efficiently as possible.

Chatbots can surprise us

To bring awareness to the water crisis, charity: water collaborated with jewelry company Lokai to build a fictional experience based in reality. Their bot (which is no longer live) was a fictional, amalgamated persona named Yeshi based on the women and children who struggle to get potable water in rural Ethiopia. Yeshi would send users little stories about her life and her fictional brother, short video clips, water trivia, and, of course, information about charity: water’s work in Africa. Interacting with the bot was a two-and-a-half minute experience, though Yeshi would go silent for 10-15 minute periods when she was “busy,” both to make the experience feel more human and more compatible with, say, working at a desk job.

While bots like this haven’t taken off on a larger scale, Yeshi offered a unique, surprising experience for those who interacted with her during her brief, simulated time with us.

Bots can also go beyond storytelling to providing services and empowering communities. WorkIt, a bot developed by Quadrant 2 and powered by IBM Watson natural language processing, enables hourly workers to find out about workplace policies from trusted and trained peer advisors. Apps like WorkIt demonstrate how bots can be powerful, engaging tools for reducing information asymmetry in various aspects of society.

Chatbots still don’t have any real competition — besides human labor

Information technology firm Gartner describes the adoption of new technology as a Hype Cycle. Essentially, when a new technology emerges, there’s a period of unrealistic expectations before the tech settles into its role in the broader ecosystem.

Gartner’s visualization of the technology Hype Cycle

This happens all the time — think blockchain, AR/VR, 3D printing, etc. But somehow, we humans are categorically unable to recognize this cycle until we’re in the Trough of Disillusionment. While chatbots haven’t taken over the communications world yet, the question is: What other technologies could overtake chatbots? The answer: There aren’t any. Lacking a competing technology, chatbots ultimately can’t lose as long as they provide more value than the alternatives, which at the current moment consist mainly of human-powered solutions which are expensive, slow, and error-prone.

Assimilation is inevitable

We chat on our phones via SMS and Messenger. We chat at work via Slack and Google Hangouts. I’ve been writing chat applications since at least 2003, but most of them wouldn’t look like a chat application because the bot itself is hidden away from the end users. Chat “winning” isn’t about convincing people to adopt a new technology — it’s about creating a useful interaction that adds value for users. We’ll only get better at that with practice.

So chat on, true believers. We may yet have dark (quiet?) days ahead of us, but ultimately, chatbots will become so pervasive that we’ll wonder how we did anything without them.

If you think you need a chatbot today, let’s talk about how to make the most of it. Get in touch with us.