Can chatbots stop annoying users? Here’s three ways to improve.

Few things unify users on a platform like LinkedIn more than poor chatbot experiences.

To wit: Posts such as this one. And, really, when’s the last time you read someone raving about how amazing a bot was (that didn’t either sell or manage a bot themselves)?

The core of the question I, our system admin for Drift, dove into in this piece on chatbots was “is there a way to make bots that don’t tick people off?”

The conclusion? Yes. But only if you adapt a customer-first mindset, which a ton of companies don’t do.

In fact, Accenture reports that 57% of CIOs and CTOs “agree that conversational bots can deliver large return on investment for minimal effort.”

That phrase “minimal effort” might be the key.

While it’s true that once bots are built, they don’t require as much optimization or maintenance as some channels (think SEO, paid ads, etc.), they do need someone to own the platform and ensure optimal experiences.

Oh, and who’s actually going to talk to customers and prospects when they want to reach a human? (That requires effort to be good a customer service, yes?)

Simply put, set-it-and-forget-it bots are only good in certain contexts, such as help articles, resources hubs, or customer service inboxes on social platforms such as Facebook.

Here’s three (+ a bonus) ways I think we can do better:

Match bots to intent. That means your homepage visitor doesn’t need a sales pitch, and someone reading a blog isn’t interesting in booking a call with sales, either.

Ask yourself “What does a visitor to this page (or subset of pages / content) want?” and then build a bot specific to that intent. Sure, this means you’ll have possibly dozens of bots, but they’ll be far more personalized to a user’s intent.

Delay the bot firing. Proactive bots that are offering additional content or a primary CTA aren’t helpful if the user hasn’t read the content yet!

Typically, you can set a bot to fire after a specified “seconds on page” or even by scroll depth. These allow the user to actually engage with the content before your bot offers more, thus providing an additive experience — instead of pushing your CTA on them before they’ve had a chance to read.

Always make it easy to reach a human. In our work, we offer marketing services. Sometimes, a web visitor may want to talk to sales — or sometimes, they may have a question about something else.

Either way, if they want to talk, we want to listen and respond! Not only does it allow you to talk to a potential customer and gain valuable insights, but you’ll ensure they feel they received a white-glove service and personal attention.

If each digital touchpoint matters in a person’s customer experience, giving them the option to reach a human when they want to absolutely puts the customer in control. And as long as they’re talking to someone who can help, they’ll have a great experience — making them more likely to buy from you!

Bonus: Turn off the ding. You know, that little chime sound that signals the bot sending its initial message.

In my testing, I’ve seen no lift from having the chime on vs. being off, but multiple folks I talked to have indicated that is annoying to them.

What about you?

What do you find annoying or helpful from bots? Are there typical patterns you see from companies leading the way vs. lagging beind?

I’d love to read about your experiences, and am always open to continue the conversation on LinkedIn!

Caleb Rule is Inbound Marketing Manager at The Pedowitz Group and is responsible for chatbots, intent data, inbound marketing, the website, and more. This piece was adapted from his blog on Dec. 22, 2021.