By next year, 85% of all customer interactions will be handled without a human agent. Yet nearly the same percentage of consumers (79%) wants human-based service.

Computers may be more capable than ever, but most customers still prefer to interact with other human beings. The key to making machines work for customer service isn’t to bypass your agents entirely; it’s to treat automation technologies as their technicians. To get agents and tech working in tandem, use tech to:

1. Connect customers with agents.

When customers reach out to your company, the last thing they want to hear is “Please listen carefully to our menu options.” Instead of using tech to greet your clients with a computer, use it to create better matches with human agents. 

Be diligent about updating your customer profiles. Invest in a contact center solution like Five9, which uses details like interests, pain points, and call history to pair customers who call in with agents. Five9 CEO Rowan Trollope describes machines as providing the mastery, while humans provide the heart that allows customer-agent connection.

2. Tackle the easy questions.

No matter what your company does, customers tend to ask the same questions over and over. Although AI shouldn’t be used for more complex inquiries or emotional customers, most customers are happy to work with bots on the small things. Eight in 10 of them take a positive view of chatbots.

Give chatbots the first crack at SMS and social media inquiries. While many bots still struggle to speak like human beings, they can write a text or respond to a direct message in a natural-sounding way. Setting up your own Facebook Messenger bot takes less than half an hour. Unless you have API developers in house, work with a company like Twilio to set up your SMS bot. 

3. Fill in missing fields.

When one of your reps is interacting with a customer, that agent should be armed with all the relevant information. What’s the customer’s address? What’s her job title? What industry does she work in? How long has she been a customer?

Your team simply doesn’t have time to identify and input all that data themselves, and asking longtime customers for it implies you haven’t been listening. Instead, turn to a data enhancement service like Cien. Using AI, Cien scrapes data from the web into a connected CRM system. If it can’t fill in a field itself, it marks the field as missing or inaccurate. For a top-down view, it reports on accuracy by individual or group. 

4. Automate follow-ups.

The single greatest complaint customers have about the companies they work with is lacking follow-up. Most customer service teams are simply stretched too thin to follow up with every client. But for customers to consider an issue resolved, they need to hear it from the business. 

Done carefully, automated follow-ups can give customers the closure they need without burdening reps. Although most CRM systems offer an automated follow-up feature, Gmail users can use a free extension to take the task off agents’ shoulders. Web-based automation systems like Zapier can send SMS follow-ups. 

Just because automation is disrupting customer service doesn’t mean that the fundamentals of customer service have changed. People want to deal with real people, and they always will. But that doesn’t mean automation technologies can’t play a role, particularly with tasks most agents are happy to hand off. Get the balance right, and everyone wins—especially the customer. 

Brad Anderson

Editor In Chief at ReadWrite

Brad is the editor overseeing contributed content at He previously worked as an editor at PayPal and Crunchbase. You can reach him at brad at