Chatbots are still frustrating customers

Image Credit: CX Today

Automated customer communication is the future of customer service. There is no swerving it.

Instead, many companies are hurtling towards it, enjoying the increasing ease of designing, developing, and delivering chatbots.

What’s more, use cases are also maturing rapidly. Many models now feed from backend integrations to pull personalized information and complete more complex transactions.

Thanks to this rising sophistication, many promising conversational AI statistics are coming to the fore, as highlighted in a thought-provoking Cyara Infographic.

One excellent example is that 58 percent of customers say chatbots have changed their customer service expectations for the better. Yet, this is only true when chatbots function as intended.

Unfortunately, this is a significant catch, as it is often not the case. Despite use cases becoming more complex, the same problems that have plagued chatbots since their inception persist.

As the previously referenced infographic suggests, many customers still experience the following chatbot frustrations:

• Unable to get answers to simple questions (80 percent)

• Redirected to an agent, and I had to repeat everything (76 percent)

• Using the chatbot was time-consuming (76 percent)

• The conversation felt impersonal (71 percent)

• The conversation lacked relevance to the actual issue (65 percent)

The good news is that quick fixes to many of these problems are readily available. After all, solutions providers now design development tools for non-technical customer care personnel.

As referenced in a recent Cyara article, Opus Research reported that “if you can play a videogame, you can build a bot.”

Yet, many businesses fail to precisely pinpoint where chatbots go wrong. And, when companies fail to spot these issues, customers find them instead.

Overcoming Chatbot Frustrations

The overarching perception that dampens attitudes towards chatbots is that they are less helpful than human agents. Such negativity often stems from a lack of quality assurance (QA).

Unfortunately, with regard to time and resources, manual testing is strenuous, and it is also confusing. Many chatbot owners will ask:

• How can we test every possible customer behaviour they may make in the future?

• How can we confidently guarantee the quality of chatbot interactions when we never honestly know what our customers may ask the bot?

Thankfully, automation testing tools can streamline the process. The trouble is that very few vendors offer these technologies across automated channels. Enter Cyara.

After its recent acquisition of Botium, Cyara now offers a codeless test automation solution that assesses the basic conversation flow of chatbot experiences.

Such capabilities are excellent for accelerating the initial training and testing of conversational AI. Yet, they also aid companies in continuously monitoring chabots and ensuring they reply as intended, post-deployment.

As a result, it becomes easier to build positive user experiences, which helps to bolster the reputation of a company and the service it provides.

Building on this point, Christoph Börner, Senior Director Digital at Cyara, said: “QA chatbot testing tools offer a blueprint to success, ensuring that conversational AI implementations and improvements deliver the expected results.”

“What’s more, as Cyara offers a complete omnichannel solution, we can even track the success of escalations and deflections to and from the channel.”

Despite these significant benefits, just three percent of global companies that offer chatbot-driven customer service (and could easily afford a QA testing platform) exploit the power of automated testing, according to Cyara.

Meanwhile, the other 97 percent must instead invest lots of time and energy in manual testing or perhaps ignore chatbot QA entirely, as the frequency of customer frustrations suggests.

Such statistics suggest that companies can do much more to qualm customer aggravations and quickly eradicate issues such as: “Using the chatbot was time-consuming,” and, “The conversation lacked relevance to my actual issue.”

Specialist testing resources are the answer, enhancing and safeguarding automated customer journeys.

Source: CX Today