Ada Helps Tile Find 291 Percent ROI

Tile sells the trackers people can attach to their keys, wallets, smartphones, remote controls, or any other object that they frequently misplace. The company’s technology, which connects to a smartphone app through Bluetooth, locates 6 million items on average for users each day.

With millions of units sold across 200 countries, a strong customer experience (CX) strategy is a must, especially during the holiday shopping season when the San Mateo, Calif., company does most of its annual business.

Traditionally Tile has been able to find enough human agents to staff its contact center during the busy season, but Justin Michaud, its senior customer support manager, figured that a chatbot would give it a competitive edge. “The bot could handle a lot of the typical questions, the simple FAQs, the troubleshooting, etc., and allow our agents to focus on the more detailed, technical things,” he says.

As he began the search for a suitable product, he recognized that “we didn’t want a solution that was too big. We wanted something that was reasonably priced for our size,” he recalls.

Tile eventually selected Ada in the summer of 2020, in part, Michaud says, because its solution integrated smoothly with the Zendesk customer service technology its was already using. The Ada chatbot went live a few months later.

“Once the contract was signed, going live with Ada was actually pretty smooth,” Michaud says. “We were able to provide them with the basic reasons that people typically contacted us, a sample of the macros that we used for those situations.”

Tile also provided Ada with 30 FAQs, which the vendor used to help design answer flows. Michaud didn’t want to overload the bot, so the company worked with Ada to develop a workflow to move callers to human agents if their queries went beyond the bot’s capabilities.

“For the most part, our products are very easy to troubleshoot,” Michaud says.

But there are other issues that require a live agent to handle. Whereas some companies would let bots handle all queries until they get stuck, Tile chose another direction.

“Our goal was for the bot to handle the typical quick FAQs that don’t require any other internal tools,” Michaud says.

Since being launched, Tile’s bot has evolved to where it is handling about 60 types of issues.

If the bot can’t handle a query, it sends the caller to an agent rather than risk giving a wrong answer that would ultimately frustrate customers.

Tile permits the bot to respond to a customer’s question only once before transferring the call to an agent. The handoff occurs in real time, a capability that was also one of Ada’s huge selling points.

In addition to seeing the bot nearly double the number of questions it can handle, Tile also within a year realized a 291 percent return on investment, highlighted by a cost savings of more than $433,000 and a reduction in agent handling time equivalent to 25,000 hours.

Based on the success it’s had so far, Tile plans to integrate the Ada bot more deeply into its operations, Michaud says. “If the bot is more deeply connected in the app, then we can pass some of the metadata back to the bot to help it with personalization.”

Looking ahead, Tile hopes to even further integrate the bot into some of its back-end applications. “If the bot can know more information about users, it can provide more relevant answers to them. Hopefully, in a year from now, the bot is connected to two more aspects of personalization and customization, and more information is available to it.” 

The Payoff

Since adding Ada's chatbot, Tile has seen the following results:

• a return on investment of 291 percent;
• cost savings of more than $433,000; and
• a reduction in handling time of more than 25,000 hours.